Evaluation of principles of motor learning

in speech and non-speech-motor learning

tasks

Ramesh Kaipa

A thesis submitted in partial fulfilment of the requirements

for the Degree of Doctor of Philosophy

University of Canterbury

November 2012

1
2
 Abstract
Principles of motor learning (PMLs) refer to a set of concepts which are considered to
facilitate the process of motor learning. PMLs can be broadly grouped into principles based
on (1) the structure of practice/treatment, and (2) the nature of feedback provided during
practice/treatment. Application of PMLs is most evident in studies involving non-speech-
motor tasks (e.g., limb movement). However, only a few studies have investigated the
application of PMLs in speech-motor tasks. Previous studies relating to speech-motor
function have highlighted two primary limitations: (1) Failure to consider whether various
PMLs contribute equally to learning in both non-speech and speech-motor tasks, (2) Failure
to consider whether PMLs can be effective in a clinical cohort in comparison to a healthy
group. The present research was designed to shed light on whether selected PMLs can indeed
facilitate learning in both non-speech and speech-motor tasks and also to examine their
efficacy in a clinical group with Parkinson’s disease (PD) in comparison to a healthy group.
Eighty healthy subjects with no history of sensory, cognitive, or neurological
abnormalities, ranging 40-80 years of age, and 16 patients with PD, ranging 58-78 years of
age, were recruited as participants for the current study. Four practice conditions and one
feedback condition were considered in the training of a speech-motor task and a non-speech-
motor task. The four practice conditions were (1) constant practice, (2) variable practice, (3)
blocked practice, and (4) random practice. The feedback was a combination of low-
frequency, knowledge of results, knowledge of performance, and delayed feedback
conditions, and was paired with each of the four practice conditions. The participants in the
clinical and non-clinical groups were required to practise a speech and a non-speech-motor
learning task. Each participant was randomly and equally assigned to one of the four practice
groups. The speech-motor task involved production of a meaningless and temporally
modified phrase, and the non-speech-motor task involved practising a 12-note musical
sequence using a portable piano keyboard.
Each participant was seen on three consecutive days: the first two days served as the
acquisition phase and the third day was the retention phase. During the acquisition phase, the
participants practised 50 trials of the speech phrase and another 50 trials of the musical tune
each day, and each session lasted for 60-90 min. Performance on the speech and non-speech
tasks was preceded by an orthographic model of the target phrase/musical sequence displayed
on a computer monitor along with an auditory model. The participants were instructed to
match their performance to the target phrase/musical sequence exactly. Feedback on

3
performance was provided after every 10th trial. The nature of practice differed among the
four practice groups. The participants returned on the third day for the retention phase and
produced 10 trials of the target phrase and another 10 trials of the musical sequence.
Feedback was not provided during or after the retention trials. These final trials were
recorded for later acoustic analyses.
The analyses focused on spatial and temporal parameters of the speech and non-speech
tasks. Spatial analysis involved evaluating the production accuracy of target phrase/tune by
calculating the percentage of phonemes/keystrokes correct (PPC/PKC). The temporal
analysis involved calculating the temporal synchrony of the participant productions (speech
phrase & tune) during the retention trials with the target phrase and tune, respectively,
through the phi correlation. The PPC/PKC and phi correlation values were subjected to a
series of mixed model ANOVAs.
In the healthy subjects, the results of the spatial learning revealed that the participants
learned the speech task better than the non-speech (keyboard) task. In terms of temporal
learning, there was no difference in learning between the speech and non-speech tasks. On an
overall note, the participants performed better on the spatial domain, rather than on the
temporal domain, indicating a spatial-temporal trade-off. Across spatial as well as temporal
learning, participants in the constant practice condition learned the speech and non-speech
tasks better than participants in the other practice conditions. Another interesting finding was
that there was an age effect, with the younger participants demonstrating superior spatial and
temporal learning to that of the older participants, except for temporal learning on the
keyboard task for which there was no difference. In contrast, the PD group showed no
significant differences on spatial or temporal learning between any of the four practice
conditions. Furthermore, although the PD patients had poorer performances than the healthy
subjects on both the speech and keyboard tasks, they showed very similar pattern of learning
across all four practice conditions to that of the healthy subjects.
The findings in the current study tend to have potential applications in speech-language
therapy, and are as follows: (1) a constant practice regime could be beneficial in developing
speech therapy protocols to treat motor-based communication disorders (e.g., dysarthria), (2)
speech therapists need to exercise caution in designing speech therapy goals incorporating
similar PMLs for younger and older adults, as the application of similar PMLs in younger and
older adults may bring about different learning outcomes, (3) and finally, it could be
beneficial for patients to practise speech tasks which would require them to focus either on
the spatial or temporal aspect, rather than focussing on both the aspects simultaneously.
4
 Acknowledgements

First and foremost, I would like to thank and express my gratitude to my primary supervisor,
Prof. Michael Robb for his invaluable guidance over the past three years. There are no words
to describe how helpful he has been since the day I arrived in Christchurch. He has not only
been a dedicated supervisor but also a wonderful mentor. I have learned a lot from him over
these three years including his dedication, writing style, critical appraisal of research
methods, and teaching style. I hope to follow his model in the coming years.

I am highly thankful to my co-supervisors, Prof. Richard Jones and Dr. Maggie Lee
Huckabee. Prof. Jones has been an excellent teacher throughout the course of my PhD
project. His attention to details, organisational style, and timely feedback has helped me to
groom my research skills to a great extent. His presence on the supervisory committee has
been my source of inspiration and motivation.

Dr. Maggie Lee Huckabee contributed to my PhD project from a unique perspective. Her
constant encouragement served as an impetus to complete my PhD project on time. Her
reasoning abilities and expertise in the area of neuroscience helped me formulate my research
questions during the initial phase of my PhD project. I thoroughly enjoyed attending her lab
group meetings every week. The lab group sessions were brain storming and helped me to
develop my research skills (and also my knowledge in statistics).

I would also like to thank all the faculty and staff (past as well as present) at the Department
of Communication Disorders for their support and encouragement during the course of my
PhD. Dr. Emily Lin was very helpful in answering my questions related to statistical
analyses. A word of special thanks goes to Ms. Sue Bradley for her assistance.

I would like to thank all my participants for their valuable time. This project would not have
been complete without their participation. I wish to acknowledge the assistance of Ms. Leslie
Livingston in helping me to contact the potential participants. Furthermore, I would like to
thank Ms. Jenny Boyer at ‘The Multiple Sclerosis and Parkinson's Society of Canterbury’ for
helping me in recruiting participants with Parkinson’s disease.

5
My fellow postgraduate students at 19 Creyke Road and NZBRI have been wonderful in
supporting me to complete my project. Their suggestions, ideas and advice were invaluable
and kept me going. I would like to thank all the staff and faculty at the University of
Canterbury (UC) who directly or indirectly helped me to complete my PhD journey at the
UC.

I admire the resilience of the people residing in the greater community of Christchurch, who
in spite of a series of devastating earthquakes kept moving forward. Even though, I had to
change my project in the midst of my PhD programme, this resilient nature of the residents of
Christchurch motivated me to get back on track. I am happy and also proud to have been
associated with the community of Christchurch.

I cannot express how thankful I am to my parents. They have made innumerable sacrifices
throughout their life to bring me up. I am grateful to them, for they introduced me to the
profession of communication disorders and encouraged me to pursue my graduate studies.
They have been and will always be there for me. My brother has always been pivotal in
advising me to choose the right direction. He has not only been my brother, but also a
wonderful friend and guru. Over the past one and half years, my sister-in-law along with my
brother has been instrumental in bringing back lighter moments in our family. I would also
like to acknowledge the ever-loving nature and prayers of my in-laws, Emmy, Rohan,
Rodney, and Thanu. I also thank all my relatives who have been my well wishers, college,
and school friends who are and were a part of my pleasant memories.

I reserve this section of acknowledgements to my better half ‘Roha’ (or ‘Chakara’ as I fondly
call her). She has taken care of me and kept me happy since the day I embarked on this PhD
journey. She has been my constant source of encouragement and support over these past three
years. I have known her as a friend, girl friend and a wife, and in each role she has proved
herself to be better and better. As we embark on our new journey, I hope to take care of her as
she took care of me during my most difficult times.

Last but not the least, I thank ‘God Almighty’ who has protected all of us and continues to
protect with his blessings.

6
 Table of Contents
Abstract.................................................................................................................................................. 3
Acknowledgements ............................................................................................................................... 5
Preface.................................................................................................................................................. 11
Abbreviations ...................................................................................................................................... 12
Chapter 1. Introduction...................................................................................................................... 15
Chapter 2. Review of Literature ........................................................................................................ 18
Motor learning ............................................................................................................................. 18
Performance during practise vs. performance during retention/transfer...................................... 18
Nature of the skill........................................................................................................................... 19
Performance characteristics of skill learning ................................................................................ 20
The stages of motor learning .......................................................................................................... 20
Theories of motor learning .......................................................................................................... 21
Closed-loop theory ......................................................................................................................... 21
The schema theory of motor learning ............................................................................................. 24
Principles of Motor Learning ...................................................................................................... 27
Application of PMLs in non-speech tasks ...................................................................................... 27
Practice condition........................................................................................................................... 27
Feedback condition ......................................................................................................................... 38
Application of PMLs in speech tasks .............................................................................................. 44
Practice condition ........................................................................................................................... 44
Feedback condition .......................................................................................................................... 50
Role of pre-practice in motor learning ....................................................................................... 52
Effect of age on motor learning ................................................................................................... 53
Spatial and temporal aspects of motor learning ........................................................................ 57
Statement of the Problem............................................................................................................. 59
Hypotheses .................................................................................................................................... 59
Chapter 3. Methods ............................................................................................................................ 63
Participants ................................................................................................................................... 63
Non-clinical group.........................................................................................................................63
Clinical group.................................................................................................................................63

Procedure ...................................................................................................................................... 66
Experiment I: Non-clinical group ................................................................................................... 67
Experiment II: Clinical group ......................................................................................................... 86

7
 Data Analyses ................................................................................................................................ 86
Spatial analysis ............................................................................................................................... 86
Temporal analysis........................................................................................................................... 87
Statistical analyses ........................................................................................................................ 91
Measurement reliability................................................................................................................... 92
Chapter 4. Results ............................................................................................................................... 94
Spatial learning ............................................................................................................................. 94
Non-clinical group.........................................................................................................................94
Clinical group ................................................................................................................................. 97
Non-clinical vs. clinical..................................................................................................................97
Summary of key findings for the spatial learning......................................................................... 100
Temporal learning ...................................................................................................................... 101
Non-clinical group......................................................................................................................101
Clinical group ............................................................................................................................... 101
Non-clinical vs. clinical ................................................................................................................ 105
Summary of the key findings for the temporal learning ............................................................... 107
Age effect ..................................................................................................................................... 108
Spatial learning ............................................................................................................................ 108
Temporal learning ........................................................................................................................ 115
Summary of key findings of age effect on spatial and temporal learning .................................... 121
NASA Task Load Index ............................................................................................................. 122
Speech task ................................................................................................................................... 122
Keyboard task ............................................................................................................................... 124
Speech vs. keyboard tasks ............................................................................................................. 124
Summary of key findings of NASA task load index......................................................................128

Chapter 5. Discussion ....................................................................................................................... 128

Spatial Learning ......................................................................................................................... 128
Non-clinical group............................................................................................................129
Speech task..............................................................................................................................129
Keyboard task..........................................................................................................................131
Speech vs. keyboard........................................................................................................132

Temporal Learning..................................................................................................................... 132

Non-clinical group ..................................................................................................................... 132
Speech task....................................................................................................................132
Keyboard task.........................................................................................................................133
Speech vs. keyboard.......................................................................................................134

Age effect .................................................................................................................................... 135

Spatial learning ............................................................................................................................ 135
Temporal Learning ....................................................................................................................... 139

8
Non-Clinical vs. Clinical group ................................................................................................. 140
Spatial Learning ........................................................................................................................... 141
Temporal learning ........................................................................................................................ 143
The non-speech vs. speech debate ............................................................................................. 146
The non-speech vs. speech debate in the context of the present study .......................................... 148
Limitations of the Study ............................................................................................................. 149
Directions for future research ................................................................................................... 152
Clinical Implications................................................................................................................... 155
Conclusion.......................................................................................................................155

References....................................................................................................................................157

APPENDICES ............................................................................................................................. 175

Appendix 1: Project Information Sheet for participants ............................................................... 176
Appendix 2: Participant consent form .......................................................................................... 180
Appendix 3: Edinburgh Handedness Inventory ............................................................................ 182

9
10
 Preface

This PhD thesis conforms to the referencing style recommended by the American
Psychological Association Publication Manual (5th ed.) and to the spelling conventions
recommended by Oxford Dictionary (http://oxforddictionaries.com/).

The research for this PhD project was carried out between September 2010 and July 2012
while the student was enrolled as a PhD candidate at the Department of Communication
Disorders in the University of Canterbury. Research expenses incurred for the purchase of
equipment (keyboard) and participants’ travel expenses were funded by the Department of
Communication Disorders.

Currently, aspects of this PhD research have resulted in three presentations at national and
international conferences
1. Kaipa, R., Robb, M. P., Huckabee, M. L., & Jones, R. (2011). Application of motor
learning principles in speech & non-speech tasks. Paper presented at the Annual
Convention of American Speech and Hearing Association, San Diego, November
17th-19th.
2. Kaipa, R., Robb, M. P., Huckabee, M. L., & Jones, R. (2012). Practice makes
perfect: is it applicable to speech motor learning? Paper presented at the Biennial
conference of New Zealand Speech language Therapy Association, Auckland, May
23rd-25th.
3. Kaipa, R., Robb, M. P., Huckabee, M. L., & Jones, R. (2012). Evaluation of motor
learning principles in healthy and clinical groups. Paper presented at the Annual
Convention of American Speech and Hearing Association, Atlanta, November 15th-
17th.

11
 Abbreviations

AD Alzheimer’s disease

ANOVA analysis of variance

AOS apraxia of speech

ASHA American Speech, Language and Hearing Association

CAS childhood apraxia of speech

CI contextual interference

CPF challenge point framework

CV consonant-vowel

EEG-EMG electroencephalography-electromyography

HD Huntington’s disease

PD Parkinson’s disease

PMLs principles of motor learning

PPC percentage of phonemes correct

PKC percentage of keystrokes correct

KR knowledge of results

KP knowledge of performance

UPDRS Unified Parkinson’s disease rating scale

min minutes

ms milliseconds

NASA National Aeronautics and Space Administration

NSOMSE Non-speech-oromotor speech exercises

s seconds

12
SAT spatial-accuracy trade-off

SLP Speech Language Pathologist

UK United Kingdom

USA United States of America

VC vowel consonant

VMT visuomotor tracking

13
14
 Chapter 1. Introduction
Speech-motor control is broadly defined as the neuronal actions that initiate and regulate
muscle contractions for speech production (Netsell, 1983). The speech-motor system refers to
the neural mechanisms used to produce speech. The efficient functioning of the speech-motor
system is affected in a sub-group of speech disorders referred to as motor-speech disorders
(MSDs) (Darley, Aronson, & Brown, 1975; Duffy, 2005). MSDs may be caused by
disruption at high levels of neural (cerebral) activity or at lower levels such as the point of
neuro-muscular junctions. MSDs include both developmental and acquired forms of
dysarthria and apraxia of speech. Individuals with MSDs represent a substantial proportion
among individuals with speech disorders (Duffy, 2005). As MSDs represent deficits in motor
control, treatment modalities focussing on aspects of motor learning/re-learning could be
useful to treat the speech deficits associated with MSDs.

Motor learning
The process of motor learning is essential for either learning new skills (e.g., a baby
learning to walk) or re-learning the lost skill(s) (e.g., an adult re-learning to walk after a
stroke). Motor learning refers to ‘a set of processes associated with practice or experience
leading to relatively permanent changes in the capacity for movement’ (Schmidt & Lee,
2005, p. 302). Often the terms ‘performance’ and ‘learning’ are used interchangeably within
the scope of motor learning, and it is essential to distinguish them. According to Magill
(2004), performance is a behaviour which can be observed and refers to the act of executing a
motor skill. Performance is not indicative of permanent acquisition of a motor skill. Learning
is a behaviour which cannot be observed but can be inferred based on a person’s
performance. Learning results in permanent acquisition of a particular motor skill. Motor
learning is usually assessed through tests of retention and transfer. Retention refers to the
persistence in the performance of an acquired motor skill, whereas transfer is indicative of
ability to perform a particular task as a result of practising another task (Schmidt & Lee,
2005).

Principles of motor learning

Even though there is no agreed standard definition of Principles of Motor learning
(PMLs), in a general context, PMLs refer to a set of guided principles thought to facilitate
learning/re-learning of motor skills when applied systematically. PMLs can be broadly
classified into two types: (1) principles pertaining to the structure of practice and,

15
(2) principles pertaining to the nature of feedback (Mass et al., 2008). Structure of practice
refers to the act of rehearsing behaviour repeatedly for the purpose of improving or mastering
it (Poole, 1991). A practice regime can be structured based on variables such as practice
amount, practice distribution, practice variability, practice schedule, source of attention, and
complexity of the practising task (Bislick et al., 2012). Nature of feedback refers to
information related to the sensation associated with the movement itself (e.g., feel, sound), as
well as information related to the result of the action with respect to the environmental goal
(Kawashima et al., 2000). Efficient feedback can be provided based on frequency, type, and
timing (Bislick et al., 2012).

Application of PMLs in non-speech-motor learning

PMLs have largely emerged from studies involving non-speech-motor tasks (e.g.,
keyboard entry tasks). Among the practice conditions, constant practice vs. variable practice
and, random practice vs. blocked practice have received considerable attention. Constant
practice refers to the practice of the same target repeatedly (e.g., practising a golf swing),
whereas variable practice targets more than one variant of a given target (e.g., practising a
golf swing over varying distances from the hole) (Shoenfelt, Synder, Maue, McDowell, &
Woolard, 2002). The benefit of variable over constant practice (for retention) has been
confirmed for a variety of tasks (e.g., Lee, Magill & Weeks, 1985; Wulf & Schmidt, 1997;
Shoenfelt et al., 2002).
Random practice refers to a practice condition in which target movements with different
motor plans are practised in such a manner that the learner is unable to predict the target for
the upcoming trial (Knock et al., 2000). Blocked practice refers to a practice condition in
which the learner practices a set of target movements and then practises another set of target
movements successively (Knock et al., 2000). Numerous studies have shown the benefits of
random over blocked practice (for retention) across a wide range of tasks (e.g., Lee & Magill,
1983; Shea & Morgan, 1979; Wright, Black, Immink, Brueckner, & Magnuson, 2004; Wulf
& Lee, 1993).
Among the feedback conditions pertaining to PMLs, feedback frequency and feedback
timing have received considerable attention. Feedback frequency refers to how often
feedback is provided during practice (Hula et al., 2008). Studies investigating frequency of
feedback have shown an advantage for low-frequency feedback (e.g., Winstein & Schmidt,
1990). Feedback timing refers to when feedback is provided relative to the completion of the
motor movement (Hula et al., 2008). Studies have shown that providing feedback

16
immediately after a task is less effective for learning than delaying it for a few seconds (e.g.,
Swinnen, Schmidt, Nicholson, & Shapiro, 1990).

Application of PMLs in speech-motor learning

Limited studies have experimentally applied and validated PMLs to speech-motor
learning. Studies examining practice conditions have shown variable practice to be beneficial
over constant practice (e.g., Adams & Page, 2000), and random practice to be beneficial over
blocked practice (e.g., Knock et al., 2000). A more recent study comparing random vs.
blocked practice conditions in children with childhood apraxia of speech (CAS) revealed both
practice conditions to be beneficial in treatment of CAS (Mass & Farinella, 2012).
In terms of feedback condition, Hula et al. (2008) investigated the effect of feedback
frequency and timing on acquisition, retention and transfer of speech skills in adults with
apraxia of speech (AOS). They found that a low frequency and delayed feedback seemed to
facilitate acquisition, retention, and transfer of speech skills.
Most of the above studies related to speech-motor learning pose two primary limitations.
First is the failure to directly compare the effects of PMLs on both speech and non-speech-
motor learning tasks within the same individual. As PMLs are largely based on studies
related to non-speech tasks (e.g., finger tapping, keyboard entry), it is essential to examine
the role of PMLs in speech-based tasks to be considered as a valid approach to the treatment
of MSDs. There have been few attempts to include PMLs on the execution of speech-motor
tasks. As there are limited studies which directly compare the effects of PMLs on speech and
non-speech-motor learning tasks, it is difficult to determine whether the PMLs found to be
effective in non-speech-motor learning are also effective for speech-motor learning. The
second limitation is that past studies have not considered the combined effects of practice and
feedback variables on speech-motor tasks. It is likely that providing optimal practise, as well
as feedback conditions might assist an individual to learn a motor task better than providing
either practise or feedback condition alone. In an attempt to address these two limitations, the
purpose of the current study was to investigate the effect of selected PMLs (practice
conditions, as well as feedback conditions) on both non-speech and speech-motor learning in
individuals with normal speech-motor control and impaired speech-motor control.

17
 Chapter 2. Review of Literature

Motor learning
Motor learning is an important psychophysiological phenomenon. It is through motor
learning we learn a variety of motor skills necessary for our daily activities. Magill (2004)
defined learning in general, and motor learning in particular, as ‘a change in the capability of
a person to perform a skill that must be inferred from a relatively permanent improvement in
performance as a result of practice or experience’ (p. 193). Two important aspects of learning
can be deduced from this definition. First, learning indicates that an individual has acquired a
new skill permanently. Second, learning cannot be observed directly, rather it has to be
inferred based on the changes in the behaviour that can be observed.

Performance during practise vs. performance during retention/transfer

In the context of motor learning, it is important to distinguish performance during practice
and performance during retention and/or transfer. Performance, in general, refers to any
behaviour that is observable. Specifically, performance refers to execution of a specific motor
skill in a specific environment (Briseno, Diaz, Romo, & Ruiz, 2010). Performance during
practice is also referred to as the acquisition phase of learning. It could be easy to gauge an
individual’s ability to learn a motor skill by observing his/her performance during the
practice regime; however, it does not provide information about the motor learning ability of
an individual. For example, observing a person hitting a baseball would imply that the
person’s performance of the skill of hitting a ball is being observed, but it does not provide
information on whether the person has learned to hit the ball correctly or not. Performance of
a motor skill during the practice regime is influenced to a large extent by performance
variables. These include factors like the alertness of the individual, the practising
environment, and the fatigue caused by practice (Magill, 2004). In summary, three important
aspects of performance during practise are: (1) performance of a motor skill during practice is
observable, (2) the effect of performing a motor skill during practice is temporary (i.e., it
does not result in learning of the skill), and (3) performance is influenced by practice
variables.
Performance during retention/transfer determines the extent of learning. Learning would
indicate that there is a permanent change in an individual’s performance as a result of
practice, and is not affected by the performance variables. Assessing learning through
retention examines the persistence of improved performance as a result of practice. The usual

18
way of administering a retention test is to have an individual perform a practised motor skill
after a certain time interval during which the individual has not practised the skill. Assessing
learning through transfer tests examines the extent to which practice on one skill generalizes
to other skills (e.g., practising the forehand shot in tennis and assessing whether the backhand
shot improves). In summary, the term ‘motor learning’ implies that: (1) learning is
permanent, (2) learning can be observed directly, (3) learning is not affected by performance
variables, and (4) learning is typically assessed by retention and/or transfer tests.

Nature of the skill

An important factor which could determine the outcome of motor learning is the nature of
the skill to be learned. Motor skills can be classified in three categories based on: (1) the
stability of the environment (open or closed), (2) precision of the movement (gross or fine),
and (3) the distinctiveness of the beginning and end points (continuous, serial, or discrete)
(Galligan, 2000; Davis, 2000). Each of the three categories has been summarised below:
Closed and open skills – This classification is based on the effect of the stability of the
environment on motor skills (Knapp, 1967). Stability would refer to whether the
environmental features are stationary (e.g., a tree) or in motion (e.g., escalator). Skills
performed in stable environment are called closed skills (e.g., lifting a pen from the table,
throwing a dart at a target). On the other hand, skills performed in a changing (unstable)
environment are called open skills (e.g., stepping onto a moving escalator, driving).
Gross and fine motor skills – Gross motor skills require large muscle groups for execution
(e.g., sitting, walking, running). Fine motor skills require relatively smaller muscle groups for
execution (e.g., writing, operating scissors) (Davis, 2000).
Continuous, serial, and discrete skills – Skills can also be classified based on their start
and end point. A skill is said to be discrete if it has well defined starting and finishing points.
(e.g., throwing a ball). Skills which have an arbitrary start or end point are called continuous
motor skills (e.g., swimming, walking). In the case of discrete motor skills, the start and end
points are determined by the performer and not by the task. When a series of discrete motor
skills are put together, it results in a serial motor skill. (e.g., bowling a cricket ball). In the
case of serial motor skill, a series of movements must be performed in a specific order to
complete the task (Galligan, 2000).

19
Performance characteristics of skill learning
Generally four performance characteristics are evident as skill learning takes place
(Magill, 2004). First, is the improvement in the performance of the skill. The second
characteristic is the development of consistency, and this implies that movement
characteristics across the multiple practice trials of the same task tend to be fairly similar. The
third characteristic is persistence; this indicates that the individual is able to demonstrate the
improved capability in performance over a longer period of time. The fourth and final
performance characteristic is adaptation. This would mean that an individual who has
demonstrated improved capability in the performance of motor skills can generalize and
adapt to a variety of other performance characteristics.

The stages of motor learning

An individual trying to learn a novel motor skill achieves proficiency in that skill only
through repeated practice. Every individual who masters a novel motor skill, typically goes
through three distinct stages of motor learning: cognitive stage, associative stage, and the
autonomous stage (Fitts & Posner, 1967). The three stages are summarized below as
described by Fitts and Posner.
Cognitive stage – In this stage, the learner spends a considerable amount in trying to
understand what needs to be done and the nature of the skill. Considerable cognitive activity
is required during this stage, as the novel learner is initially unsure of what needs to be done.
The attentional demand for the movement production is very high. The performance during
this stage is highly variable in nature, the movements are gross and large number of errors is
observed. Even though, the learner is aware that the movements are incorrect, he/she is not
sure of how to correct the movements. The learner consciously controls majority of the
movements. During this stage, the learner is heavily dependent on feedback in the form of
verbal instructions, demonstrations, and guidance. The gains in performance are largest
during this stage as the learner constantly explores the strategy that improves the
performance. There is a tendency for learner to become easily frustrated if success is not
achieved quickly.
Associative stage – Once the basic movement pattern is achieved, the learner enters the
associative stage of motor learning. During this stage, the performance is less variable and is
more consistent. Less cognitive activity is required, and the learner is dependent on
proprioceptive feedback rather than visual or auditory feedback. The errors are reduced and
the movements become more refined. The learner develops an ability to detect his/her errors

20
during the movement, even though this ability is not perfect. The attentional demand for
movement production reduces. During this stage, some parts of the movement are controlled
consciously, whereas some are automatically performed. The learner begins to concentrate on
perfecting the skill. This stage may last between few days to months.
Autonomous stage – After an extensive period of practice, the learner enters the
autonomous stage of learning. This stage reflects the highest level of proficiency, and not all
learners reach this stage. Here, the performance becomes consistent and reliable. The
movements are automatic and do not require any cognitive effort or attention. The
movements are effortless, and are free of errors most of the time. Usually it takes years of
practise to reach this stage of learning. The learner develops an ability to detect his/her errors
during the movements and tends to correct those errors. To retain the skill at this stage, the
skill must be repeatedly practised.
In summary, motor learning cannot be strictly delineated into these three stages, as the
process of motor learning reflects a continuum. However, these stages of motor learning best
explain the trajectory of learning of a novel motor skill. The learner gradually proceeds from
one stage to another instead of an abrupt change.

Theories of motor learning

Within the context of motor learning, theories serve to explain the exact process of motor
learning. There are two important theories which have had a significant impact on
understanding motor learning and are discussed below.

Closed-loop theory
Adams (1971) was the proponent of the closed-loop theory of motor learning. He
developed this theory through a series of experiments involving slow lever-operating tasks.
Adams suggested that the principles of performance and learning applicable to these
experiments could be generalized to other motor movements as well. This theory emphasized
the importance of feedback to learn a motor task and suggested that motor learning proceeds
through the gradual refinement of perceptual-motor feedback loops (hence the name closed-
loop theory). When performing a novel motor task, the initial movements are crude and are
not effective to achieve the intended outcome. During further practice trials, the perceptual
feedback associated with the motor movements provide information about the particular
location of the limbs in space, and whether the movements were able to achieve the target
motor goal. This information provided by perceptual feedback is referred to as the
“perceptual trace”. With each subsequent practice trial, the perceptual trace guides the

21
individual to produce motor movements which resemble the correct motor goal (also known
as the correct trace). Eventually, through a combination of movements guided by the
perceptual trace, the individual achieves the correct motor goal. The basic premise of the
closed-loop theory is that feedback guides an individual to perform tasks more accurately.
When people learning new tasks are told explicitly about their performance, they tend to do
better than people who do not receive such feedback. Thus, the basic function of feedback is
to guide the novice learner to achieve the intended motor goal through subsequent practice.
The process involved in learning a novel motor skill as explained by closed-loop theory is
shown in Figure 1 (a-c). During the early learning phase (Figure 1a), the individual produces
an equal number of correct as well incorrect movements. The movements tend to be
inconsistent and highly variable during this phase. During the subsequent phase (Figure 1b),
the individual guided by the verbal feedback learns to produce movements which begin
resembling the correct motor trace, but some of the motor movements still continue to be
inaccurate. During the third phase (Figure 1c), the individual starts producing more
movements which are closer to the correct trace, and in this process the number of incorrect
movements are reduced to a substantial extent.
One of the important implications of the closed-loop theory is the error detection
capabilities developed by the learners during the course of practice. Each time a movement is
made, the learner compares the accuracy of his/her movements to the target motor goal
through the feedback provided by the perceptual trace. This difference between the
performed movement and the target movement is referred to as the ‘error’. If there is a large
error during the initial stages of practice, the learner attempts to reduce these errors during
subsequent practice trials by producing motor movements which are close to the target goal.
This error detection capability eventually helps an individual to learn a novel motor task
efficiently. The closed-loop theory has been criticized for not accounting for two major
aspects of movement. First, the closed-loop theory is based on slow, lever positioning tasks.
It does not intend to explain the motor control/learning of rapid action movements. Research
has shown that feedback acts too slowly for learning rapid tasks like throwing and ball-
striking (Henry and Rogers, 1960; Keele, 1968). The second criticism is that the theory does
not account for generality (Schmidt, 1975b). For example, an action can be performed in
many non-identical yet similar ways.

22
 (a)

(b)

(c)

Figure 1. An illustration of Adam’s closed-loop theory (from Schmidt & Lee, 2005, p.411). As repetitions
accumulate, the perceptual trace starts approximating the correct trace. With many repetitions, the shape of the
distribution becomes more peaked at the mode. Panels a-c depict the early, middle, and late phase of motor
learning, respectively as described by Adam.

23
The schema theory of motor learning
Dissatisfied with Adam’s closed-loop theory, Schmidt (1975b) formulated the schema
theory. This theory is based on an open-loop process which was not accounted by the closed-
loop theory. Adams proposed that when an individual trying to learn a new movement makes
an error, the feedback loop guides him/her to correct his/her error(s) during the subsequent
practice trials. On the other hand, Schmidt and his colleagues provided evidence that it takes
about 120-200 ms for the whole process of sensory error detection, and initiating appropriate
corrections in response to those errors (Schmidt & White, 1972; Schmidt & Wrisberg, 1973).
These researchers also noted that some of the sensory channels like proprioception operate at
a speed of about 110 ms (time taken to respond to an external stimulus). Based on this
evidence, Schmidt argued that even though feedback is important for motor learning, the
feedback loop advocated by the closed-loop theory cannot account for learning rapid motor
movements. Owing to this shortcoming, Schmidt proposed that motor movements are
performed based on a set of pre-defined motor commands called generalized motor programs
(GMPs) which are not dependent upon feedback loop. GMPs are assumed to be a set of pre-
structured commands designed to execute a range of motor movements if response
specifications are provided. These response specifications are parameters that can be varied
before the movement begins that enable the motor program to be performed at a different
speed or a different force. For example, a motor program for hitting a ball can be modified to
be performed slow or fast based on the response specifications (Schmidt, 1975b; Schmidt &
Lee, 2005).
According to the schema theory, an individual is able to generate novel motor movements
based on the notion of “schema”. In order to generate novel motor movements, it is essential
that the schema stores information about four important aspects related to the motor
movement: (1) the initial conditions, (2) the response specifications of the motor program, (3)
the sensory consequences of the outcome, and (4) outcome of the movement. Each of these
aspects is summarised below as described by Schmidt (1975b).
Initial conditions - This would refer to the environmental conditions in which the
individual performs the motor movement. The initial conditions comprises (though not
limited to) the individual’s pre-muscular state, and the information received from various
receptors about the surrounding environment (e.g., proprioceptive information about the
position of limbs in space, visual, and auditory information) (Keele, 1968; Pew, 1974). Once
the movement is executed, the initial conditions used to plan the movement are stored.

24
 Response specifications - The motor program required for the generation of motor
commands is rather general, so the commands are fine-tuned and refined by response
specifications. Response specifications (parameters) govern the manner in which the motor
commands are executed. For example, response specifications are responsible for changing
aspects like speed, force, and direction of the motor movements. Once the movement is
executed, these specifications associated with the movement are stored for further use.
Sensory consequences - These refer to the information (feedback) received from the eyes,
ears, and proprioceptors after the execution of the movement. This sensory information
received after the execution of movement is also stored for further use to develop appropriate
schemas related to motor movement.
Response outcome - This is the fourth aspect of information stored after the movement and
it provides information about the success of the response in relation to the original intended
outcome. This is commonly referred to as knowledge of results (KR), such as “you kicked the
ball 2 feet away from the centre of the goal post”. Thus, this KR provides information as to
how successful the response outcome was and is stored after the completion of the
movement.
Schema formation - The initial conditions, response specifications, sensory consequences,
and response outcomes are stored together after a movement is performed. When the same
movement is performed repeatedly, then the individual begins to draw information about the
relationship among these four sources of information. The strength of this relationship among
the four sources of information increases with each subsequent movement, and this relation is
the schema for that particular movement. Thus the knowledge of particular movements is
stored as ‘motor schemas’ by individuals.
When an individual is required to perform a motor movement for which he already has a
motor plan, then the movement is initiated with information received from two types of
schema: the recall schema and the recognition schema (Schmidt, 1975b). The recall schema
encodes the relationship between initial outcome, response specifications, and the intended
outcome. When the recall schema is supplied with information about the initial conditions
and the intended outcome, then it formulates the appropriate response specification necessary
to generate the specific motor movement. On the other hand, recognition schema encodes the
relation between the initial outcome, sensory consequences, and the intended outcome. When
the recognition schema is supplied information about the initial condition, and the intended
outcome then it computes the appropriate sensory consequences associated with a specific
motor movement (Schmidt & Lee, 2005). The recognition schema generates two types of
25
expected sensory consequences: (1) the expected proprioceptive feedback, and (2) the
expected exteroceptive feedback which consists of visual and auditory feedback. If there is a
mismatch between the expected sensory consequences and the actual sensory consequences,
then this represents an error signal which is used to update the recall schema. Thus, the recall
and recognition schemas work in unison to ensure the smooth ongoing execution of motor
movements (Schmidt, 2003).
Even though the schema theory explains the generation of novel motor movements
through a generalized motor program (GMP), a drawback of this theory is that it does not
explain how the GMP is formed in the first place. This theory also does not take into account
as to how the rules about response specification and sensory consequences are formulated
(Schmidt & Lee, 2005).
In summary, closed-loop theory (Adams, 1971) and schema theory (Schmidt, 1975b)
explain the process of motor learning from different perspectives. Closed-loop theory
mentions that an individual trying to learn a new motor skill makes a number of errors to
begin with. However, through constant refinement of the perceptual feedback loop during the
practice regime, the individual is able to learn the correct movement pattern of the motor
skill. The schema theory was basically formulated to account for some of the shortcomings of
the closed loop theory which were not effective to explain the entire aspects of motor
learning. Schema theory contradicts the notion of motor learning through perceptual feedback
as some of our sensory channels (like proprioception) operate at an extremely rapid pace
which is much faster than the time taken to receive the perceptual feedback. Hence, Schmidt
proposed that motor movements are learned and executed based on a set of pre-structured
motor commands – GMPs – and GMPs are refined by response specifications (parameters).
Critical examination of closed loop and schema theories suggest that the contribution from
both the theories could be relevant in motor learning. Even though Schmidt mentioned the
role of GMP in motor learning, no explanation was provided as to how the GMP is developed
in the first place. It is likely that an individual trying to learn a new motor skill might initially
rely on perceptual feedback to develop a prototype of the correct movement pattern of the
practising skill. This prototype can serve as a comparator model to guide the learner to detect
his/her errors during each practice trial and refine his/her movement patterns to resemble the
original movement pattern. Once the prototype has been well developed, it can in turn
facilitate the formation of a GMP corresponding to the original movement pattern and
thereby reducing the learner’s dependency on perceptual feedback. Thus, the role of both
theories needs to be acknowledged within the scope of motor learning.
26
Principles of Motor Learning
Previous research has established that motor learning is indicative of permanent
acquisition of new skills (Poole, 1991). However, the extent of learning a motor skill is
guided by structured application of certain PMLs. PMLs refer to a set of principles intended
to facilitate the process of motor learning (Magill, 2004). PMLs have their roots from the
closed-loop (Adams, 1971) and schema theories (Schmidt, 1975b), and since then have been
used extensively to study the behavioural aspects of motor learning and also in clinical
settings. Most of our knowledge regarding PMLs is derived from experiments pertaining to
limb-based tasks (Lee & Magill, 1983; Lai & Shea, 1998). The application of PMLs in
learning novel speech tasks and to treat speech disorders has been emerging in recent years
(Mass et al., 2008; Bislick et al., 2012). The PMLs are divided based on the structure of
practice and nature of feedback.

Application of PMLs in non-speech tasks

There have been a number of studies pertaining to the application of motor learning in
non-speech tasks and these studies have helped to evaluate the efficacy of PMLs in
learning/re-learning various motor skills. A review of several studies in regards to practice
and feedback conditions is provided below.

Practice condition
A primary reason for practising a skill is to attain mastery or to perfect the skill (Gentile,
1972). Practice, in a general context of motor learning would refer to the repeated rehearsal
of a motor behaviour. The practice conditions consist of (1) amount of practice, (2) practice
distribution, (3) practice variability, (4) practice schedule, (5) attentional focus and (6)
holistic practice (Mass et al., 2008). The various practice conditions used in learning/re-
learning of non-speech as well as speech-motor skills are shown in Table 1. Each of these
conditions is described below as applies in non-speech tasks.
Practice amount - The amount of practice an individual devotes to a skill is critical for
learning a motor skill (Magill, 2004). Practice amount refers to the time spent practising
movements (Mass et al., 2008). To learn a motor skill, it is essential that some amount of
practice has to be undertaken by the learner. The amount of practice needed depends on the
complexity of the task and how much expertise is needed. Some research has even proposed
that specific motor skill expertise is gained through accumulating an average of 10,000
practice hours (i.e., typically 10 years) (Ericsson, Krampe, & Tesch-Romer, 1993). It is
essential to estimate the amount of practice needed by an individual to learn a particular

27
Table 1. Practice conditions with appropriate examples for application in non-speech and speech tasks (adapted from Mass et al., 2008).

Options Relationship to non-speech (learning Relationship to speech (learning to say

Practice (bolded options are more to play tennis) “aeroplane”)
condition desirable for motor
learning)
Amount Small vs. Large Practising a serve 5 times vs. 50 Practising to say the word “aeroplane”
times 5 times vs. 50 times

Distribution Massed vs. Distributed Practising 50 serves in 10 min vs. 50 Practising to say the word
serves in 25 min “aeroplane” 50 times in 5 min vs. 50
times in 10 min
Variability Constant vs. Variable Practising the serve in the same spot Practising to say the word
vs. practising in different spots “aeroplane” at a constant rate of
speech vs. saying it at different rates

Schedule Random vs. Blocked Practising forehand and backhand Practising the words “aeroplane” and
shots randomly vs. practising “ship” randomly vs. practising
forehand shot 20 times and then “aeroplane” 20 times and then
proceeding to backhand shot proceeding to “ship”

Focus Internal vs. External Focusing on the hand grip vs. Focusing on the lips vs. trying to hear
watching the movement of the the word while saying it
racket

Holistic Simple vs. Complex Practising a straight drive vs. a spin Practising individual syllables in the
shot word “aeroplane” vs. the whole word

28
motor skill. One of the common problems encountered in determining the amount of practice
is that it might result in over-learning or under-learning of the task. Magill (2004) suggests
that in order to achieve expertise, more practice is better than less.
The context of ‘overlearning’ has received considerable attention in motor learning over
the past years. Overlearning refers to the continuation of practice even after attaining mastery
over the skills. Past research has proven overlearning to be beneficial in learning novel skills
(Dirskell, Willis, & Cooper, 1992; Bromage & Mayer, 1986). Dirskell et al. reviewed 15
studies from 1929 to 1982 investigating the role of overlearning on motor learning. The 15
studies involved 3,771 participants. The participants in some of the studies practised only
physical tasks (e.g., balancing a stabilometer), whereas participants in other studies practised
only cognitive tasks (e.g., remembering verbal information). They found that the degree of
overlearning in these 15 studies ranged from 0% overlearning to 200% overlearning, and the
benefits gained from overlearning were reduced by one-half after 19 days of practice. Based
on their findings, the researchers suggested four important aspects related to overlearning: (1)
overlearning is beneficial in terms of enhancing the retention of the tasks, (2) overlearning is
effective in learning both physical and cognitive based tasks, (3) the retention benefits are
directly proportional to the degree of overlearning (e.g., 150% overlearning enhances results
in more learning than 50% overlearning), (4) the benefits of overlearning may disappear at
longer retention intervals.
The effect of overlearning on long-term retention was studied by Rohrer, Taylor, Pashler,
Wixted, and Cepeda (2005). The researchers recruited 218 college students as participants for
two experiments. The first experiment involved 130 students who studied 10 city-country
pairs (e.g., Moscow-Russia), one group of students practised this task four times more (high
learners) than the other group (low learners). The participants came back for the retention
tests at one and nine weeks after the initial practice. Results revealed that, after one week, the
high learners recalled much more than the low learners. However, after nine weeks the
retention of the high learners reduced by about two-thirds (from 70% at week 1 to 24% at
week 9), while the retention of the low learners reduced by less than half during the same
period (from 31% at week 1 to 17% at week 9). In the second experiment, 88 students studied
a word-definition task (e.g., cess-tax). Similarly, the learners were divided into low learners
and high learners. The participants returned for retention tests at intervals of one week and

29
four weeks. The results revealed that the high learners had a significant advantage over the
low learners after one week. However, this advantage disappeared after four weeks.
Specifically, the retention of the high learners declined by about two-thirds (from 64% at
week 1 to 22% at week 4), whereas the retention of the low learners declined by about one-
half (from 38% at week 1 to 18% at week 4). The results of this study were in agreement with
the findings of Dirskell et al. (1992), and suggests that overlearning could be beneficial for
short-term retention but not for long-term retention.
With regards to a clinical population, Kwakkel (2009) systematically reviewed studies
pertaining to intensive rehabilitation after stroke and found a dose-response relationship; that
is, patients who received more practice showed improved functional outcome in comparison
to patients who received less practice. Kwakkel reviewed 20 randomized control trials
involving 2686 patients either in sub-acute, post-acute, and chronic stage after stroke. All the
patients in the studies received either physical or occupational therapy to improve their
activities of daily living. Among these 2686 patients, some of the patients received more
intensive rehabilitation compared to other patients. On average, the patients in the intensive
rehabilitation groups received about 959 minutes more rehabilitation than the control groups.
Results revealed that the patients who received higher intensive rehabilitation improved
significantly more than the control groups.
Dirskell et al. (1992) suggested that overlearning provides more opportunity for attaining
initial proficiency in learning a task. However, the more important aspect is that practice
beyond this initial proficiency allows the learner to receiver further feedback about the
correctness of the response and this feedback helps in longer retention of the task.
Overlearning (or a large amount of practice) helps in retention but the exact duration of the
beneficial effects of overlearning remains to be investigated.
In summary, there is clear indication from past studies that more practice or overlearning
is beneficial in learning motor skills. However, the term ‘overlearning’ has been used
arbitrarily in past studies (Dirskell et al., 1992; Rohrer et al., 2005). There is no clear
specification as to how much practice (in terms of practice trials or number of hours)
constitutes overlearning. Further research could possibly investigate the threshold (in terms of
number of practice trials or hours spent in practice) which clearly delineates learning from
overlearning of a range of motor skills.

30
 Practice distribution - Practice distribution refers to how a given practice regime is spaced
across time. This ranges from massed to distributed practice. In massed practice, an
individual practises a certain number of trials within a shorter time frame with no rest or a
very short rest interval between the practice sessions or trials (Schmidt, 1991). In a
distributed practice, the individual practises the same number of trials across a longer period
of time and the rest interval between the practice sessions or trials is also relatively longer
(Burdick, 1977).
Past studies have indicated that distribution practice tends to be beneficial over massed
practice in learning skills (Baddely & Longman, 1978; Rohrer & Taylor, 2006). Baddely and
Longman compared the effect of massed and distributed learning on a keyboard task. The
training time limitations were a total of 60 hours and five days each week. The practice
sessions were distributed in four different ways. Two groups practised one hour in each
session. Among these two groups, one group practised only one session each day, thus
prolonging the total practice session for 12 weeks; whereas, the other group practised two
sessions each day, resulting in a practice regime which lasted for 6 weeks. Two other groups
practised two hours each session. One of these groups practised only one session each day,
resulting in a practise regime for 6 weeks, and the other group practised two sessions each
day, thereby reducing the practise span to 3 weeks. The most distributed practice regime
required the participants to practise for 12 weeks, and the most massed practice regime
required the participants to practise for 3 weeks. The outcome measures were based on the
number of hours required to learn the keyboard task, and typing speed. Results indicated that
the most distributed practice group required the least amount of time to learn the keyboard
task and had the fastest typing speed. The most massed practice group required the longest
time to learn the task and was the slowest in terms of the typing speed.
Rohrer and Taylor (2006) investigated the benefit of massed vs. distributed practice on
solving mathematical problems. The participants were 216 college students who were
randomly assigned to massed and distributed practice groups. The students in the massed
practice group solved 10 mathematical problems in one single session, and students in the
distributed practice group solved the 10 problems in two separate sessions separated by one
week. Retention tests after one week revealed that there was no difference between the two
groups. However, a retention test after four weeks revealed that students in the distributed
practice group were more efficient in solving the problems compared to the students in the
massed practice group.

31
 There are three possible reasons suggested by Magill (2004) to explain the beneficial
effects of distributed practice. First, individuals involved in massed practice tire easily, and
fatigue negatively influences learning. Second, the continuous nature of practice involved in
massed practice reduces the cognitive resources of the learner if the practice continues
beyond a certain critical amount. The third reason pertains to the memory consolidation
process. Memory consolidation facilitates long-term storage, and for memory consolidation
to happen it is essential that there is an adequate rest interval between practice sessions.
Distributing the practice across days facilitates memory consolidation than massing the
practice within a day or two (Brashers-Kug, Shadmehr, & Bizzi, 1996). Thus, distributed
practice seems to be more beneficial than massed practice in learning new motor skills.

Practice variability - Practice variability involves practising the different movement

variations of a motor skill. A practice situation can involve an individual practising only one
variant of a skill, which is referred to as the constant practice. A practice session can also
involve an individual practising more than one variation in the dimensions of a skill, which is
referred to as the variable practice. In the case of variable practice, there is a scaling of the
motor skill (i.e. variation in the dimensions of the motor skill).
Research has revealed that variable practice tends to benefit learning over constant
practice for a variety of tasks. (e.g., Shea & Kohl, 1991; Shoenfelt et al., 2002; Wulf &
Schmidt, 1997). Shoenfelt et al. compared the effects of constant and variable practice on
shooting a basketball. The researchers found that that the constant as well as variable practice
groups improved during the acquisition phase. However, the variable practice group
demonstrated significantly better performance than the constant practice group on a retention
test after two weeks. Kerr and Booth (1978) compared the beneficial effects of constant vs.
variable practice on learning a tossing skill. Thirty-six children were recruited to learn a bean
bag tossing skill at one or more targets for a period of 12 weeks. One group of participants
practised tossing the bean bag at a target three feet away (constant practice), whereas the
other group of participants practised tossing the bean bag at two targets which were two and
four feet away, respectively (variable practice). Both groups of participants were required to
toss the bean bag at a target three feet away as a part of the post-test. The results revealed that
the variable practice group was significantly better on the post-test than the constant practice
group.

32
 The beneficial effect of variable practice is attributed to the “Elaboration Hypothesis”
(Shea & Morgan, 1979). These researchers suggest that the effect is due to the elaboration of
the memory representations of the skill variations. In the case of variable practice, the learner
can compare and contrast the skill variations which make them distinct from one another and
thus help in better retention of the skill. In the case of constant practice, the individual
practises the same variation of the skill thereby not giving an opportunity to compare and
contrast the various skill variations. This could account for the decreased beneficial effects
offered by constant practice as suggested by past studies.
There are studies which have disproved the beneficial effects of variable practice over
constant practice in learning motor skills (Dick, Beth, Shankle, Dick-Muehlke, Cotman, &
Kean, 1996; Breslin, Hodges, Steenson, & Williams, 2012). Breslin et al. compared constant
vs. variable practice in learning to shoot a basketball. Ten students in the constant practice
group practised 300 trials of basketball shooting from a constant distance of 15 feet.
Alternatively, 10 students in a variable practice group practised 300 trials of basketball
shooting across five different distances. The practice took place over two consecutive days,
and the retention test took place on the third day. Results revealed that the constant practice
group performed better than the variable practice group. Dick et al. also compared constant
vs. variable practice in learning a tossing skill. Twenty-four healthy adults and 28 patients
with Alzheimer’s disease (AD) participated in the study. The participants were required to
learn tossing a bean bag at an archery type target, and practised the task for a total of 10
weeks with two practice sessions each week. During each practice session, the participants
practised 32 trials of the tossing task. Participants in the constant practice condition practised
the task at a constant distance, whereas participants in the variable practice condition
practised at four different distances. Retention tests were conducted one week and one month
after training. The results revealed that healthy participants were benefited by constant as
well as variable practice conditions. On the contrary, participants with AD were benefitted
only by constant practice. A likely explanation for the better performance of the AD
participants in the constant practice condition is that practising multiple variations of a skill
could have overburdened their cognitive system in comparison to practising a single variation
of the skill. It is possible that constant practice could be more beneficial than variable practice
to learn motor skills in clinical populations like individuals with AD.
In summary, studies comparing constant vs. variable practice have found equivocal
results. It remains unclear as to whether constant or variable practice tends to be more
beneficial in learning/re-learning motor skills. Further research across a range of motor skills
33
will be helpful to determine whether variability of practice is indeed beneficial in motor
learning.

Practice schedule - Practice schedule refers to the order in which the practice stimulus is
presented to the learner. Practice tasks can either be scheduled in a random manner or a
blocked manner (Schmidt & Lee, 2005). In blocked practice, the practice trials of one
particular task are performed together, uninterrupted by practice on any other sequences. For
example, in the case of learning to play tennis, it would involve practising 10 trials of a
forehand serve, followed by practising another 10 trials of a backhand serve. However, in
random practice, the practice trials are intermixed and the upcoming practice trials are not
predictable. Referring to the tennis example, practising multiple trials of forehand and
backhand serves in an unpredictable and random manner within the same practice session.
The advantage of random over blocked practice has been proven in a number of studies
(Shea & Morgan, 1979; Shea & Wright, 1991; Wright, 1991). The benefit offered by the
random practice has been mainly attributed to the ‘contextual interference’ effect. Contextual
interference (CI) is a learning phenomenon wherein the interference caused due to practising
different tasks within the same practice session proves to be beneficial (Magill & Hall, 1990).
The concept of CI was first mentioned by Battig (1972) in a verbal learning task (paired-word
associations) for what he initially referred to as ‘intra-task interference’. Shea and Morgan
(1979) were among the first researchers to demonstrate the advantage of CI effect in learning
limb-based motor tasks. The researchers recruited 72 right handed students as participants for
the study. The participants were required to knock down three barriers in three different
sequences. The participants were randomly assigned to either a high CI group (random
practice) or a low CI group (blocked practice). The participants practised the task for a total
of 54 trials divided in three sets of 18 trials (for the three different sequences, respectively).
Participants in the blocked practice group practised one sequence of trials before proceeding
to the next sequence. Participants in the random practice group practised all three sequences
in an unpredictable manner during each practice set of 18 trials. Retention tests were
conducted after a 10 min delay and a 10 day delay under random and blocked practice
sequences. Results of the retention tests revealed that participants in the random practice
group had fewer sequence errors, and demonstrated faster reaction and movement time.
Sherwood (1996) also found random practice to facilitate motor learning. Twenty-four
college students were required to learn a rapid lever reversal movement so that the reversal
point was 20 , 40 , 60 , or 80 . The participants were assigned to either random or blocked

34
practice groups. All the participants practised 90 trials of the task. Retention tests
immediately after the acquisition phase and after 24 hours revealed that participants in the
random practice group showed more spatial accuracy in comparison to the blocked practice
group.
However, there are studies which suggest that the beneficial effects of random practice
cannot be generalized to all motor tasks (Brady, 2008; Maslovat, Chua, Lee, & Franks, 2004;
Meira & Tani, 2001; French, Rink & Werner, 1990). Maslovat et al. (2004) compared the
effect of random vs. blocked practice on learning a bimanual coordination task. A bimanual
task requires manipulation from both the hands (Guiard, 1987). Thirty right-handed
participants were assigned either to a blocked practice group, random practice group, or a
control group. The participants in random and blocked practice groups practised two
bimanual coordination tasks in a random and blocked practice schedule, respectively.
Participants in the control group practised a single bimanual coordination task only. All the
participants performed a total of 200 acquisition trials over two consecutive days. Retention
tests were conducted immediately following the second day, and after one week. The results
revealed that the random practice group demonstrated better performance than the blocked
practice group. However, neither the random nor blocked practice groups demonstrated better
performance than the control group, suggesting that the use of a random practice schedule
could be beneficial for learning only one task.
French, Rink, and Werner (1990) compared the benefits of random vs. blocked practice in
learning three volleyball skills among high-school students. The participants were required to
learn the forearm pass, the set, and the overhead serve (i.e., the basic arm moves in a
volleyball game). The participants were assigned to either a random practice group, blocked
practice group, or a random-blocked practice group. Retention results revealed that even
though there was significant improvement in all three groups, there were no differences
between groups.
With regards to a clinical population, Lin, Sullivan, Wu, Kantak, and Winstein (2007)
compared random vs. blocked practice conditions in learning a movement task. Twenty
healthy adults and 20 adults with mild PD served as participants. The participants were
required to operate a lever and move it horizontally at a specific speed and distance to learn a
goal movement task. The goal movement task was displayed on the computer screen before
each trial which the participants were required to replicate. Three versions of the movement
task were used. Participants in the blocked practice condition practised the three movements
in a blocked sequence for a total of 135 trials, whereas participants in the random practice
35
condition practised the three movements in a random order for a total of 135 trials. The
experiment lasted for two consecutive days. The first day was allotted for the practice phase
(acquisition phase) and the second day was the retention phase. Results of the retention test
revealed that the healthy participants in the random practice condition performed better than
the participants in the blocked practice condition. However, the results were the opposite for
the participants in the clinical group. Thus, the above studies suggest that random practice
may not be the ideal practice schedule, especially when considering a learning task for
clinical population.
In summary, past studies have revealed mixed findings with respect to the beneficial
effects of blocked vs. random practice in motor learning. It was long thought that CI offered
by random practice might benefit motor learning (Shea & Morgan, 1979; Magill & Hall,
1990). However, studies over the recent years have disproved this notion (Brady, 2008;
Maslovat et al., 2004). Further research is required to identify the best practice schedule that
would facilitate motor learning across a range of motor skills and clinical populations.

Attentional focus - Attentional focus refers to the source of attention during the process of
motor skill learning (Wulf, 2007). The focus of attention can be either an internal source or
an external source. An internal focus of attention refers to the attention directed by the learner
towards his/her own body movements. An external focus of attention refers to directing
attention to the role of learner’s body movements on the surrounding environment (Vance,
Wulf, Tollner, McNevin, & Mercer, 2004). Usually an external focus of attention is
considered to be more beneficial in learning motor skills rather than an internal focus of
attention (e.g., Wulf, HÖb, & Prinz, 1998; Wulf & McNevin, 2003). For example, Wulf et al.
compared the effects of internal vs. external focus of attention on learning a skiing task.
Thirty-three participants were recruited to learn slalom-type movements on a ski-simulator.
The participants were randomly assigned to one of the three groups (internal focus group,
external focus group, and a control group). The internal focus group received instructions to
focus on their feet while performing the task. The external focus group received instructions
to focus on the wheels of the platform located directly under the feet, while the control group
was given no focus instructions. The participants practised the task on two consecutive days
and a retention test was conducted on the third day. The results revealed that the external
focus group demonstrated better learning than the internal focus and control group. There was
no difference in learning between the internal focus and the control group.

36
 The benefits of external attentional focus have been attributed to a ‘constrained action
hypothesis’ (McNevin, Shea, & Wulf, 2003; Wulf, McNevin, & Shea, 2001). According to
this hypothesis, when individuals are asked to focus on their body movements (internal
focus), they tend to constrain body movements, which serves to disrupt automatic control
processes. When individuals are asked to focus on the effect of the movement (external
focus), this manner of attention facilitates automatic processes to control the movement
resulting in effective learning. In summary, it is generally agreed that an external focus of
attention tends to be more beneficial than internal focus in motor skill learning.

Holistic practice - A motor skill can be either practised in whole or in part (Park, Wilde, &
Shea, 2004). In general, part practice is considered to be simpler in nature in comparison to
whole practice which is considered to more complex. The concept of whole vs. part practice
in learning a motor skill has been debated since the early half of the 20 th century (Barton,
1921; Knapp & Dixon, 1952; Wickstrom, 1958). Naylor and Briggs (1963) hypothesized that
the influence of whole vs. part practice would depend on two factors: (1) skill complexity,
and (2) skill organization. The complexity of a skill would refer to the number of parts in the
skill as well as the attention demands of the skill (Magill, 2004). Wulf and Shea (2002)
considered a task to be complex if it could not be mastered in a single session and had several
degrees of freedom. Degrees of freedom refer to the number of independent elements of a
movement system. For example, a lever which can be pushed forward or backward has only
two degrees of freedom. A task was considered to be simple if it could be mastered in a
single session and had only one degree of freedom. Organization of a skill would refer to the
extent of relationship among the subcomponents of a skill. A skill is said to have a high
organization if the subcomponents are interdependent on one another. A skill with low
organization will consist of subcomponents which are relatively independent of one another
(Coker, 2009). The Naylor and Briggs hypothesis has provided certain guidelines to
determine the effectiveness of part vs. whole practice in learning a motor skill. A whole skill
practice is recommended if the skill is low in complexity and high in organization. A part
practise approach is recommended if the skill has a high complexity and low organization.
Park et al. (2004) compared part-whole vs. whole practice on learning a movement
sequence task. The researchers randomly assigned 18 university students to either a part-
whole practice group or a whole-practice group. Participants in both the groups were required
to learn a 16 movement sequence using a lever. Participants in the part-whole practice group
practised only the first 8 elements on the first day (100 repetitions of the first 8-elements) and

37
all 16 elements on the second day of practice (100 repetitions of all the 16 elements). On the
other hand, the whole-practice group had to practise all 16 elements on both days (100
repetitions of the 16-element sequence per day). On transfer tests during which the first and
second 8 elements were tested separately, the participants in the part-whole practice group
revealed better performance than the whole practice group, especially on the second 8-
elements .
Dean, Kovacs, and Shea (2008) compared the transfer from smaller spatial movement
sequence to a larger sequence and larger sequence to a smaller sequence. Twenty-eight
college students participated in the study. The participants had to either operate a lever
through a bigger movement sequence (targets space at 20 , 40 , 60 , and 80 ) or through a
smaller movement sequence (20 , 26.7 , 33.3 , and 40 ). Transfer from bigger to smaller
movement sequence was more effective than the other way around. Thus, this study provides
support for the whole practice approach.
With regards to a clinical population, Nettlebeck and Kirby (1976) used part or whole-task
methods to train mild mentally retarded workers to thread a sewing machine. The researchers
found that the part-practice approach helped the participants to learn the sewing task better
than the whole practice approach.
In summary, there is no clear indication directing the use of part practice or whole practice
approach to learn a motor skill. On the other hand, a part practise approach is recommended
if the skill has a high complexity and low organization. Caution should be exercised in
generalizing these guidelines to learn a variety of motor skills. It is likely that the
effectiveness of part practice or whole practice depends on the complexity of the skill being
practised. Further research is required to determine the role of part practice vs. whole practice
in learning simple as well as complex skills.

Feedback condition
Feedback refers to the information individuals receive about their performance of a motor
skill. This information can be provided either during or after the performance (Wulf,
Chiviacowsky, Schiller, & Ávila, 2010). In other words, feedback pertains to performance-
related information. When an individual performs a motor task, two types of feedback are
available to the individual. The first type of feedback gained by an individual through his/her
sensory channels (e.g., vision, audition, proprioception) is referred to as “inherent” or “task-
intrinsic” feedback. A major extent of motor control and learning is achieved by information
received through our sensory channels (Schmidt & Lee, 2005). Proprioception, vision and

38
audition are generally regarded as the main sources of feedback during the process of motor
learning (Saunder & Knill, 2003; Perkell et al., 2000; Haith, Mial, & Vijayakumar, 2008).
The ability to sense body position and movement is known commonly as proprioception
(Grey, 2010). Through proprioceptive knowledge, we are able to sense the position of our
body and limbs in space without having to look at them. Proprioception includes the senses of
movement, vibration, position, deep pain, and equilibrium (Webb & Adler, 2008). Thus,
proprioception integrates information from other systems like somatosensory, vestibular and
visual systems. The sensory receptors of the proprioception called as the proprioceptors, are
located in the muscles and joints throughout the body. During movement, sensory signals
from the proprioceptors are conveyed to the spinal cord and higher cortical centres via the
afferent fibres (sensory fibres), and provides information about the location of our limbs in
the surrounding environment. This information conveyed to the central nervous system is
termed as the proprioceptive feedback, which is used to regulate activity in the neuronal
systems generating the commands to muscles (Grey, 2010).
Vision provides information about the performed motor movements in the environment,
and this guides our subsequent motor behaviour. Two visual systems have been implicated in
visual stimulus processing and feedback: focal vision and ambient vision (Trevarthen, 1968),
and they are briefly described. Focal vision is responsible for processing images in the central
part of the visual field, and is affected by decreasing levels of illumination. Focal vision tell
us ‘what the object is’. Ambient vision is responsible for processing images in the entire
visual field, and is not affected by decreasing levels of illumination. Ambient vision tells us
‘where the object is’. The role of vision in motor learning has been documented as early as
1934 (Melcher, 1934). Saunders and Knill (2005) mentioned that continuous visual feedback
of the hand is essential in learning fast reaching movements. Laguna (2008) described the
importance of visual observation in developing memory representation of the practising task,
which in turn facilitates motor learning.
Auditory feedback has also proved to be essential in motor learning, especially with
regards to speech-motor learning (Perkell et al., 2000). The researchers describe an internal
auditory model which is used to learn novel speech sounds. This internal model is a
representation of the articulatory configurations associated with various sounds produced in
the vocal tract. The importance of auditory feedback in re-learning speech sounds across
various clinical conditions has also been emphasized by previous studies (Ménard et al.,
2007; Kaipa, Robb, Beirne, & Allison, 2012).

39
 The second type of feedback received from an external source in addition to the intrinsic
feedback is referred to as “extrinsic” or “augmented” feedback (e.g., verbal instructions
provided by an instructor to a student who is learning gymnastics). Augmented feedback is
further divided into three categories based on: (1) the type of feedback, (2) the feedback
frequency, and (3) the timing of feedback. All the three feedback categories can interact
among one another as shown by previous studies (Adams & Page, 2000; Hula et al., 2008).
The various feedback conditions used in learning/re-learning of non-speech as well as
speech-motor skills are shown in Table 2.

Feedback type - There are two types of augmented feedback: (1) knowledge of results
(KR), and (2) knowledge of performance (KP) (Schmidt & Lee, 2005). KR refers to
externally presented information about the outcome of a movement and is provided verbally
(Winstein, 1991) (i.e., whether the learner achieved the goal of the performance or not). For
example, a golf instructor advising the learner that the ball missed the hole. In contrast, KP
provides information about the movement pattern made by the learner that led to the outcome
(Gentile, 1972). For example, a golf instructor advising the student that the ball missed the
hole because he/she did not perform the backswing adequately before the downswing was
made. In addition to providing the KP verbally, it can also be provided visually through video
replay showing the learner’s performance. There are limited studies comparing the efficacy
of KR vs. KP in learning motor skills and the results have been equivocal. In some instances,
providing KR might be redundant with the inherent feedback (Weeks & Kordus, 1998). For
example, if an archer does not hit the target, the intrinsic feedback (in the form of visual
feedback) tells him/her that the outcome was not favourable. In such an instance, providing
KR might be redundant. However, if the outcome of the performance cannot be determined
by the learner, then providing KR might prove beneficial. For example, while performing
some types of the motor skills (like gymnastics), the learner might not be able to determine
the outcome of the performance, and in such cases verbal KR could be useful.
Kernodle and Carlton (1992) compared the effects of KR and KP (in the form of videotape
relays and verbal statements) in an experiment which required the participants to throw a soft,
spongy ball as far as possible with the non-dominant arm. Results revealed that KP facilitated
learning better than KR. Another experiment conducted by Zubiaur, Ona, and Delagado
(1999) comparing KR vs. KP revealed similar results. In this experiment, students with no
prior volleyball experience practised overhead serve. KP provided information about the most
important error which needed attention, whereas KR provided information about the ball’s

40
 Table 2. Feedback conditions with appropriate examples for application in non-speech and
speech tasks.

Feedback Options Relationship to non-speech

Conditions (bolded options are (learning to play tennis) Relationship to speech (learning
more desirable for to say “aeroplane”)
motor learning)

Type Knowledge of How did I swing the racket? How did the lips move? vs. Was
performance vs. vs. did the ball land on the it said correctly?
Knowledge of results correct spot?

Frequency High vs. Low Feedback after every shot vs. Feedback after every attempt vs.
feedback after every 20 shots feedback after every 20 attempts

Timing Immediate vs. Delayed Providing feedback Providing feedback immediately

immediately after the serve after production vs. delaying by
vs. delaying by 10 s 10 s

41
spatial precision, flight and rotation. Results revealed that KP helped the students to learn the
overhead serve better than KR. Magill (2004) summarised that KR tends to be beneficial in
motor learning when: (1) learners use KR to compare with the inherent feedback about their
performance, and (2) learners are unable to determine the outcome of their performance
based on the inherent feedback. Whereas, KP might be beneficial when: (1) skills must be
performed according to certain movement characteristics (e.g., gymnastics), (2) KR is
redundant with the inherent feedback, and (3) skills with complex coordination movement
need to be improved. In summary, there have been equivocal findings with regards to the
beneficial effects of KP vs. KR in learning motor skills. It is likely that the effect of KP vs.
KR is dependent on the type of task being learned to a large extent. Further research is
required to determine the beneficial effects of KP vs. KR across a range of motor skills.

Feedback Frequency - Feedback frequency refers to how often feedback is presented

during the practice regime. The feedback frequency ranges from high (e.g., after every single
practice trial) to low (e.g., after every 20 practice trials). It was long thought that high
frequency feedback is beneficial during motor learning (Bilodeau, Bilodeau, & Schmusky,
1959). However, beginning with the seminal work of Salmoni, Schmidt, and Walter (1984),
studies have proved the other way around. The phenomenon of reversal effect with respect to
feedback frequency has been demonstrated in past studies (Winstein & Schmidt, 1990;
Vickers, 1994). In the case of reversal effect, high feedback frequency facilitates acquisition
but degrades learning.
For example, Weinstein and Schmidt (1990) compared the effects of high vs. low frequency
in learning a lever positioning task. Participants had to practise a complex movement pattern
using a lever on a tabletop. Participants practised the task for two consecutive days. One
group of participants received feedback after every single practice trial (100% feedback),
whereas the other group received feedback only for 50% of the trials. The retention test after
one day revealed that the group which received feedback on 50% of the trials performed
much better than the group which received 100% feedback. The detrimental effect of high
frequency feedback is attributed to the guidance hypothesis (Salmoni, Schmidt, & Walter,
1984; Winstein & Schmidt, 1990). According to this hypothesis, when learners receive
feedback after every practice trial (100% feedback), they become highly dependent on the
feedback to perform the motor skill. When feedback is no longer available to the learners, it
results in degrading their performance.

42
 However, this reversal effect has not been demonstrated in some studies (Wulf, Shea &
Matschiner, 1998; Marschall, Bund, & Wiemeyer, 2007). For example, Wulf et al. studied the
effect of high vs. low frequency feedback on learning a complex ski simulation task.
Twenty-seven participants practised the production of slalom-type movement on a ski
simulator for two days. The participants were randomly assigned to either a high feedback
frequency group (100%) or a low feedback frequency (50%) or a control group who received
no feedback. The outcome parameters were force onset and movement amplitude. A retention
test was performed on the third day. The retention test results revealed that the high
frequency group demonstrated the best performance, the low frequency feedback
demonstrated intermediate performance, and the control group showed the least performance.
Thus, the researchers suggested that the reversal effect of feedback frequency might not be
applicable in learning complex tasks.
In summary, low feedback frequency, as well as high feedback frequency, has been
demonstrated to be useful in learning a range of motor skills. It is likely that a high frequency
feedback schedule could be useful in learning complex motor skills, whereas low frequency
feedback could be useful in learning simple tasks.

Feedback Timing - Feedback timing refers to the time period in which feedback is
provided after the completion of the task. The timing of feedback can be either immediate or
delayed. The beneficial effect of delayed feedback has been demonstrated by past studies
(Swinnen, Schmidt, Nicholson, & Shapiro, 1990; Guadagnoli & Kohl, 2001). For example,
Swinnen et al. instructed the participants to learn a lever operating task to achieve a specific
movement-time goal. The participants received three types of feedback: immediately after
completing the task, or, 3.2 s, or 8 s after completing the task. Results revealed that providing
feedback immediately after the task completion had a negative influence on learning the lever
task. A reason attributed to the negative effect of immediate feedback is that it blocks the
learners’ own analysis of inherent feedback which is essential for the development of error-
detection capabilities (Swinnen et al., 1990; Guadagnoli & Kohl, 2001). Thus, past studies
suggest that delayed feedback seems to be more beneficial than immediate feedback.
In summary, with regards to practice condition, there is sufficient evidence to support
large amount of practice, distributed practice, and external focus of attention to be more
beneficial than less practice, massed practice, and internal focus of attention, respectively.
There have been equivocal findings in the case of constant vs. variable practice, random vs.
blocked practice, and part vs. whole practice. Within the feedback condition, past studies

43
favour delayed feedback over immediate feedback, and there has been mixed findings with
respect to KP vs. KR, and high vs. low feedback frequency.

Application of PMLs in speech tasks

The research related to the application of PMLs in learning speech tasks and to treat
speech disorders has been gradually increasing since the last decade (Adams & Page, 2000;
Knock et al. 2000; Adams, Page, & Jog, 2002; Steinhauer & Grayhack, 2000; Mass et al.
2002; Mass et al. 2008; Mass & Farinella, 2012; Bislick et al. 2012). A summary of some of
these past studies is provided in Table 3. There have been limited studies investigating the
application of PMLs in speech-motor learning and are heavily drawn from the limb-based
tasks. As the beneficial effect of a number of PMLs is yet to be determined in speech-motor
learning, only the PMLs which have proved to be beneficial in learning speech tasks or
treating speech disorders are discussed below.

Practice condition
Practice amount – There has been a recent spur of interest in investigating the amount of
practice required to learn/re-learn various speech and language tasks. A recent issue of the
Journal of International Speech Language Pathology was dedicated to the discussion of
amount of practice required for treatment of various speech disorders (Baker, 2012a, 2012b;
Packman & Onslow, 2012; To, Law, & Cheung, 2012; Roy, 2012; Yoder, Fey, & Warren,
2012; Enderby, 2012; Manes & Robin, 2012). The above studies address practice amount in
terms of frequency of intervention for various speech and language disorders like stuttering,
aphasia, motor-speech disorders, voice disorders, reading disorder, and speech sound
disorders. Warren, Fey, & Yoder (2007) mentioned that the intensity of an intervention
programme is based on a number of factors like: (1) what is being carried out in intervention
sessions (active ingredients), (2) the number of times the active ingredients occur in a therapy
session (dose), (3) number of intervention sessions per unit time (e.g., per day/per week), (4)
the duration of individual intervention sessions, (5) and the length of the entire intervention
programme (total intervention duration). Thus, it might be logical to deduce that the intensity
of an intervention programme might vary based on all the above factors. In addition, most of
the above studies seem to suggest that it is always not possible to prescribe an ideal amount
of practise in speech language pathology, as the amount of practise is dependent on
extraneous variables like what is being carried out in intervention, type of disorder, cognitive
status of the client, family environment of the client, financial status of the client.

44
 Table 3. Summary of various studies related to application of PMLs in speech-motor learning

Study Participants Practice task Practice condition investigated Outcome

Adams & Page 40 healthy participants Temporal learning of a Constant vs. Variable, Random Variable, random and low frequency
(2000) speech utterance vs. Blocked, and Low frequency feedback were beneficial in learning
vs. High frequency feedback the utterance

Knock et al. (2000) Two adults with Learning CV and VC Random vs. Blocked Random practice was beneficial over
aphasia and AOS syllables (treatment study) blocked

Adams, Page & Jog 18 participants with PD Temporal learning of a Low frequency vs. High Low frequency was beneficial over
(2002) speech utterance frequency feedback high frequency

Steinhauer & 30 healthy participants Vowel nasalization task Low frequency vs. High Low frequency facilitated the vowel
Grayhack (2000) frequency feedback nasalization task

Mass et al. (2002) Two adults with Learning monosyllable words Simple vs. Complex stimuli Complex stimuli practise was more
aphasia and AOS (treatment study) practice beneficial

Mass et al. (2008) Provided information regarding the

(review paper) application of PMLs in speech related
tasks and treatment

Wong, Ma & Yiu 21 participants with Sentence reading task Constant, random, and blocked No significant difference between the
(2011) vocal hyperfunction practice three conditions

Edeal & Neumann Two children with Treatment of consonant High production frequency vs. High frequency production was
(2011) AOS targets Low production frequency beneficial over low frequency

Maas & Farinella Four children with Treatment of various speech Random vs. blocked practice Mixed findings
(2012) AOS stimuli like monosyllables,
bisyllable words

Bislick et al. (2012) Provided information regarding the

(review paper) scientific rigour of the past studies
related to PMLs in SLP.

45
Study Participants Practice task Practice condition investigated Outcome

To, Law, & Cheung 102 SLPs Survey study A survey related to treatment The treatment intensity provided by SLPs varied
(2012) intensity provided for speech depending on their work settings and in most cases it
sound disorders was investigated was not sufficient.

Baker (2012a) Practice amount Provided information about ways of determining

(discussion paper) frequency intervention in SLP practise.

Yoder, Fey, & Practice amount The authors reviewed one of their past studies and
Warren (2012a) suggested that spacing the treatment sessions can have
(commentary paper) an impact on treatment intensity

Roy (2012) Practice amount Discussed regarding the harmful effects involved in
(commentary paper) excess practise of vocal exercises.

Enderby (2012) Practice amount Amount of therapy provided to clients must be based
(commentary paper) on factors like impairment, psychosocial aspects.

Manes & Robin Practice amount Provided information regarding different practice and
(2012) feedback conditions which can have an impact on
(commentary paper) deciding the practise amount (e.g., practice variability
can affect the practice amount).

Packman & Onslow Practice amount Provided information about Lidcombe programme,
(2012) and mentioned that since Lidcombe is mainly parent-
(commentary paper) driven, it is difficult to prescribe the exact practice
amount in such cases.

Baker (2012b) Practice amount The author reviewed all the above studies related to
(discussion paper) practice amount and concluded that recommending
ideal practice amount is dependent on extraneous
factors like the client’s impairment, psychosocial
status, financial status.

46
The literature pertaining to the practice amount in non-speech tasks recommend
‘overlearning’ to be beneficial in motor learning. However, the same recommendation might
be harmful for some patients with certain speech disorders. For example, Roy (2012)
mentioned that overdose of voice therapy might result in vocal fold tissue damage rather than
benefitting the patient. In summary, the results of recent studies would indicate that that there
is no universal prescription for the amount of practise in speech language pathology, and the
recommended amount of practise should be based on the factors discussed above. This is in
contrast with studies considering the amount of practise in non-speech learning tasks
(Dirskell et al., 1992; Bromage & Mayer, 1986).

Practice variability – Adams and Page (2000) compared constant vs. variable practice in a
group of 40 healthy participants. This experiment also investigated the effects of practice
schedule and feedback frequency on learning a novel speech utterance recruiting the same
cohort of participants. The participants were assigned to one of four different groups. One
group of participants practised 50 trials of the utterance “Buy Bobby a Poppy” with the target
duration of 2.4 s (constant practice), and the other group practised 50 trials of the same
utterance with the target durations of 2.4 s and 3.6 s (variable practice). The performance
feedback was provided after every practice trial for participants in both groups through
graphing the utterance durations. The participants underwent a retention test two days after
the acquisition phase and produced the target utterance without further practise. The outcome
measure was the absolute error (AE), which was determined by calculating the absolute
difference between the target utterance duration and the participants’ utterance durations.
Each participant’s AE score was based on the last five trials of the retention phase. The AE
score was obtained for the 2.4 s target duration. The results revealed that both groups
demonstrated similar performance during the acquisition phase, but the retention test results
indicated that the variable practice group had significantly lower AE in comparison to the
constant practice group. The results of this study suggest that variable practice is beneficial in
learning speech tasks which is in close agreement with some of the studies related to non-
speech-motor learning (Shea & Kohl, 1991; Shoenfelt et al., 2002; Wulf & Schmidt, 1997).

Practice schedule – The findings of the studies which have compared random vs. blocked
practice conditions have been equivocal. Adams and Page (2000) compared random vs.
blocked practice conditions on learning the same utterance task “Buy Bobby a Poppy” (as
noted above). The same participants and experimental protocol were used to carry out this
experiment. The retention results two days after the training revealed that the random practice

47
group had significantly lower AE in comparison to the blocked practice group. The results of
this experiment suggests that random practice is favourable in learning speech tasks and is in
agreement with some of the findings related to non-speech tasks (Shea & Morgan, 1979;
Shea & Wright, 1991; Wright, 1991).
In regards to a clinical population, Knock et al. (2000) compared random vs. blocked
practice in treating speech deficits in two adult males who presented with AOS, as well as
aphasia. A single-subject alternating treatment design was chosen for the study, so that each
participant served as his own control. The first participant underwent two phases of
treatment. In phase 1, the production of CV syllables (e.g., /pa/, /ba/, /ta/) were treated, and in
phase 2 the production of VC syllables (e.g., /ap/, /ab/, /at/) were treated. The second
participant underwent only one phase of the treatment. The treatment stimuli for the second
participant included six CVC words (e.g., cat, tap, vase). Each treatment phase comprised of
12 treatment sessions and each treatment session consisted of a blocked practice condition
and a random practice condition. The order of the practice conditions was counterbalanced
across the 12 treatment sessions. The practice stimuli used during the treatment sessions were
also tested during the retention phase. Results of the retention test revealed that stimuli
trained using random practice had greater retention than the stimuli trained using blocked
practice. The retention effects were more pronounced after four weeks of treatment. This
trend was noticed in both participants. The results of this study suggest that random practice
may be more beneficial over blocked practice in treating certain speech disorders.
Recently, Mass and Farinella (2012) compared the effect of random vs. blocked practice
condition in treating CAS. Four children with CAS participated in the study. A two-phase
alternating treatment design with multiple baselines across behaviours and a
withdrawal/maintenance component was used as the experimental design. The Dynamic
Temporal and Tactile Cueing method (DTTC) was used in the treatment. The DTTC method
uses PMLs for speech practice and feedback delivery, and also incorporates auditory, visual,
and tactile cueing by using a specific hierarchy of temporal delay between stimulus delivery
and response. Each treatment session contained random and blocked practice conditions and
the conditions were counterbalanced across sessions. The treatment targets varied for the four
participants depending on the severity of CAS. The treatment lasted for four weeks and each
treatment condition (using blocked or random practice) was followed by a two-week
maintenance interval to measure retention. In addition to retention, transfer was also assessed
on untreated but related words. The results were mixed, with two participants benefitted by
blocked practice, one participant by random practice, and another participant did not show an

48
improvement in either condition. The findings of the Mass and Farinella are not in agreement
with the findings of Knock et al. (2000). Mass and Farinella attributed the difference in
findings to the age of the participants in both the studies. Specifically, Knock et al. recruited
adult participants as opposed to Mass and Farinella who recruited children.
Wong, Ma, and Yiu (2011) compared random, blocked, and constant practice in learning
of relaxed phonation in patients with vocal hyperfunction. Twenty-one patients with
hyperfunctional voice problems were randomly assigned to one of the three above mentioned
practice conditions. Participants in the constant practice condition practised reading sentence
stimuli with four Chinese characters. Participants in the random practice condition practised
reading sentence stimuli varying in length from two to five characters in a random manner,
and participants in the blocked practice condition read the sentence stimuli in increasing
length starting from two to five characters. Surface EMG feedback from the orofacial and
thyohyoid region was provided to the participants after reading every two sentence stimuli.
The participants underwent eight sessions of training which lasted for four weeks. A retention
test after one week of training revealed that considerable voice motor learning was
demonstrated by participants in all of the practice conditions, and there was no significant
difference in learning between the three practice groups. Similar to the findings of non-
speech-motor learning (Brady, 2008; Maslovat, Chua, Lee, & Franks, 2004; Meira & Tani,
2001; French, Rink & Werner, 1990), recent studies related to speech-motor learning have
also revealed equivocal findings with regards to the beneficial effects of random vs. blocked
practice conditions.
Holistic practice - Mass et al. (2002) compared the effect of part (simple) vs. whole
(complex) stimuli in treating speech deficits associated with AOS. A withdrawal design along
with a multiple baseline design across behaviours was used in treating speech deficits in two
patients with combined AOS and aphasia. The researchers used the framework of part-whole
syllable structure to define the stimuli complexity. The stimuli used for the treatment were
non-words. A whole syllable structure with three-element s-clusters comprised the complex
condition (e.g., spleem), whereas the part syllable structure (singletons) comprised the simple
condition (e.g., leem). Both patients were subjected to two counterbalanced treatment phases
(a simple stimuli phase and a complex stimuli phase). The transfer effect of treatment speech
targets was investigated by using untrained real word stimuli which were related to the
treatment targets. For the first participant, treatment using complex (whole) stimuli resulted
in overall improvement in production of simple and complex real as well as nonwords. The
same effect was also observed for the treatment carried out using simple (part) stimuli but to

49
a lesser extent. For the second participant, the treatment using simple as well as complex
stimuli resulted in an improvement in the production of simple real and nonwords but not in
the production of complex real and nonwords. The results of this study suggest that speech-
motor learning may be most beneficial when using complex stimuli compared to simple
stimuli as part of the practice condition.

Feedback condition
Feedback frequency – Studies comparing high vs. low frequency feedback in learning
speech tasks and to treat speech disorders have found that low frequency feedback tends to be
more beneficial over high frequency feedback (Adams & Page, 2000; Adams, Page, & Jog,
2002; Hula et al., 2008; Mass et al., 2012), which is similar to the findings of some of the
studies related to non-speech tasks (Salmoni et al., 1984; Weinstein & Schmidt, 1990).
Adams and Page (2000) compared low frequency vs. high frequency feedback in learning
the utterance task “Buy Bobby a Poppy”. Participants in the high frequency group practised
50 trials of the speech utterance with target duration of 2.4 s and received feedback after very
single trial. Participants in the low frequency group also practised 50 trials of the same task,
but received feedback after every five trials. Participants were given feedback about their
performance through graphing the utterance duration values. Retention test results two days
post-training revealed that participants in the low frequency feedback group performed better
than participants in the high frequency group.
Hula et al. (2008) examined low vs. high frequency feedback in treating speech deficits
associated with AOS. Four participants with AOS participated in this experiment. A single-
subject alternating treatment design was used, so that each participant received both the
treatment condition (high frequency feedback and low frequency feedback) and also served
as his/her control. Each participant received two phases of treatment. In phase 1, CV
combinations beginning with fricatives (e.g., /fa/, /vu/) were treated using high frequency
feedback, and CV combinations beginning with plosives (e.g., /pa/, /ba/) were treated using
low frequency feedback. In phase 2, this arrangement was reversed. The order of sessions
during the treatment was counterbalanced across the participants. There was also a four-week
maintenance phase following the treatment phase. Weekly probes administered throughout
the 16 weeks (treatment and maintenance phases) served to assess the retention when
treatment was removed. Learning was also assessed through transfer using untrained stimuli
probes. The results revealed that low frequency feedback enhanced retention in two
participants, and transfer effects were seen in only one participant. The results reveal that

50
some of the participants were able to benefit from treatment using low frequency feedback.
The main reason for the difference in treatment outcome could have been due to differences
in the severity of AOS among the four participants.
In a recent study, Mass et al. (2012) investigated the effect of high frequency vs. low
frequency feedback frequency in treating speech deficits in four children with CAS. The
children ranged in age from 5;4 (years;months) to 8;4. The treatment targets were chosen
depending on each child’s speech and language status. The DTTC was used in the treatment.
An alternating treatment design with multiple baselines across behaviours was used. Each
child received high frequency feedback as well as low frequency feedback within a single
session, with the order of feedback conditions counterbalanced across sessions. The post-
treatment results revealed mixed findings, with two children benefitted from low frequency
feedback, one child benefitted from high frequency feedback (to a small extent), and the other
child not benefitting from either condition. The researchers suggested that although reduced
feedback might benefit children with CAS in general, it may vary with the child’s age and the
severity of the apraxia.

Feedback timing – There are few studies comparing immediate and delayed feedback
in speech-related tasks. Hula et al. (2008) conducted a second experiment which compared
immediate vs. delayed feedback. The two participants who completed the feedback frequency
experiment also took part in this experiment. There was a gap of one week between both
experiments. Feedback was provided for both the participants after every trial either
immediately (immediate feedback) or after a delay of 5.0 s (delayed feedback). Retention and
transfer results revealed that only one participant demonstrated treatment gains. Based on the
findings of this experiment, it is difficult to assess the benefits of delayed feedback in speech-
motor learning. Further research using a larger cohort of participants is required to compare
the benefits of immediate vs. delayed feedback in learning speech tasks.
In summary, studies investigating various practice and feedback conditions in speech-
related tasks are limited in number in comparison to the studies related to non-speech tasks.
With regards to practice condition, variable practice, and complex stimuli practice are
considered beneficial over constant practice and simple stimuli practice, respectively. There
have been equivocal findings with respect to random vs. blocked practice condition. In
feedback condition, low frequency feedback, and delayed feedback are considered to be
beneficial over high frequency feedback, and immediate feedback, respectively. The efficacy

51
of other practice and feedback conditions remains to be investigated in speech-motor
learning.
As speech is also a motor activity, it is plausible that the PMLs applicable to limb-based
tasks might also be applicable for speech-related tasks. However, previous research has
shown that speech and non-speech activities differ based on the degree of movement
coordination (Grimme, Fuchs, Perrier, Schoner, 2011), neural resources (Smith, 2006), and
cognitive demands (Grimme et al., 2011). So it is unclear whether the PMLs found to be
effective for non-speech motor learning would also be effective for speech motor learning

Role of pre-practice in motor learning

Pre-practice helps an individual to prepare for the upcoming practice sessions/trials
(Edwin, Karyll, Lise, & Gary, 1981). The role of pre-practice in motor learning has been
gaining considerable attention in recent years (Schmidt & Lee, 2005; Murray, McCabe, &
Ballard, 2011; Bricker-Katz, McCabe, Lincoln, & Ballard, 2011). The aim of a pre-practice
session should be to assure that the person understands the importance of the task to be
learned and performs the task in the correct manner. The three important aspects of pre-
practice are: (1) to build adequate motivation for learning the task, (2) providing correct
information about the task to be learned, and (3) modelling the task to be learned (Sherwood
& Lee, 2003), and they are described below.

Motivation for learning

It is essential that an individual remains motivated while learning a motor task. If the
motivation level decreases, then the individual might not find the task engaging, resulting in
decreased learning outcomes. Two ways to motivate an individual who is learning a task are
by (1) making the task seem important, and (2) setting appropriate and achievable goal (Mass
et al., 2008). When an individual is trying to learn a task, he or she should be informed about
the importance of learning the task and the benefits involved in learning the task. For
example, a person with dysarthria who is attending speech therapy, should be informed about
the importance of speech therapy and the associated benefits of treatment activities. Setting
specific goals which are achievable is also very important in motivating an individual who is
learning a new task. It is recommended to avoid “do your best goals” as they tend to be very
general rather than being specific (Mass et al., 2008). For example, in the case of a person
with cluttering, it is better to set the goal of reducing the rate of speech to a specific number
of words per minute rather than telling the person “to do your best”.

52
Verbal information
Orienting a novice learner about the task through effective verbal instructions has proven
to be an important precursor to motor learning (Schmidt & Lee, 2005). Instructions can be
provided regarding the task to be performed, how to perform the task, what are the outcome
measures to be achieved, and the possible error detection strategies after the task is
performed. Care should be exercised that the exact amount of information is provided to the
learner through verbal instructions. Providing too much or too little instructions can prove to
be detrimental in learning the task.

Modelling
Another important aspect of pre-practice is modelling. The task which is to be learned can
be modelled in many ways. An effective way of modelling is to demonstrate the skill directly
to the learners so that they can observe the specific steps involved in executing the task. The
task can also be modelled through videotapes or photographs of skilled performers (Svinin,
Riken, Goncharenko, Hosoe, & Kanou, 2007).

Effect of age on motor learning

Aging is an inevitable biological process experienced by every living organism. Aging is
accompanied by a gradual decrease in physiological output (Partridge & Mangel, 1999). Mild
irreversible changes in the functioning of most of the body organs start by the third or fourth
decade of our lives with progressive deterioration in aging (Boss & Seegmiller, 1981). Some
of the age-related systemic changes include decreased cardiac output, hypertension,
decreased lung volume, decreased renal functioning (Boss & Seegmiller, 1981), and decrease
in the number of neurons in the cortex (Brody, 1955). Aging is also accompanied by changes
in the sensory system (Shimokata & Kuzuya, 1995). Old age results in a decrease in visual
acuity (Shinomori, Schefrin, Werner, 2001), hearing acuity (Tremblay, Piskosz, & Souza,
2003), taste, and smell perception (Schmall, 1993).
A hallmark characteristic of old age is the decline in cognitive-motor performance. Past
research has revealed that aging is characterized by impairments in motor (Ketcham &
Stelmach, 2001) as well as cognitive functioning (Dixon & Raz, 2000; Salthouse, 1985; Cook
& Woollacott, 2000). Old people tend to perform motor tasks with less precision and more
slowly in comparison to their younger counterparts (Voelcker-Rehage, 2008). Fozard,
Verryyssen, Reynolds, Hancokc, and Quilter (1994) showed that the reaction to an auditory
stimulus decreases by 0.5 ms between 20 and 96 years. Myerson, Hale, Hirschman, Hansen
and Christiansen (1989) measured reaction time of the older and younger participants in three

53
different cognitive-motor tasks (letter classification, abstract matching-to-sample, and choice
reaction time). The researchers found that the performance of the younger group was better
than the older group in the cognitive-motor tasks. The researchers suggested that age has a
direct effect of physical slowing of movements. A number of models have been developed to
explain the decline of motor skill with age.
A popular model which explains the decreased and slower motor performance among
elderly individuals is the “information loss model” (Myerson, Hale, Wagstaff, Poon, &
Smith, 1990). According to this model, response planning requires several processing stages,
and a certain amount of time is dedicated to each stage. More time is required to plan a
response, if there is a loss of information at any particular stage. Aging is accompanied by
increased information loss at each processing stage, and hence more time is spent in planning
the movement responses. Another model explains the decreased motor output among the
elderly based on changes in “attitudes and preferences” (Verhoff, Reuman, & Feld, 1984). A
critical assumption is that elderly people approach a task differently than their younger
counterparts and are more resistant to novel and unfamiliar tasks, thus resulting in a
decreased motor output.
Crossman and Szafran (1956), and others (Welford, 1985) attributed the decreased and
slower motor performance of the elderly group to brain-based changes. This “neural noise
model” model mentions that as people age, there is increased random activity in the brain
referred to as noise. In the case of complex learning tasks, the noise activity is further
heightened. Due to this increased noise, additional time is required to integrate incoming
information from the external world involved in producing a response.
The “reduced working memory model” is well supported by previous studies (Kester,
Benjamin, Castel, & Craik, 2002; Jost, Bryck, Vogel, & Mayr, 2011). The reduced working
memory in elderly individuals is attributed to the inhibitory deficit hypothesis. According to
this hypothesis, older individuals are unable to inhibit interference from task-irrelevant
information; this irrelevant information interferes with the essential information and reduces
the memory capacity of the aged individuals (Hasher & Zacks 1988; Hasher, Zacks, &
Rahhal, 1999; Zacks, Hasher, & Li, 2000).
Despite the changes in motor ability and learning that accompany old age, previous
research has well established that elderly individuals are capable of learning/re-learning
motor skills (Seidler 2006; Ketcham & Stelmach, 2001). However, the extent and the style of
learning might differ between elderly and younger individuals (e.g., Strickgold & Walker,
2005). Walker, Brakefield, Morgan, Hobson, and Strickgold (2002) demonstrated that elderly

54
individuals are capable of retaining and improving their skills after a period of delay. Various
studies related to motor learning of non-speech and speech-motor tasks in elderly individuals
are mentioned below.

Non-speech motor learning

Research investigating the effect of age on non-speech-motor learning tasks suggests that
older individuals tend to demonstrate decreased motor learning in terms of precision and
speed in comparison to the younger individuals (Baron & Menich, 1985; Seidler, 2006;
Breitenstein, Daum, & Schugens, 1996). Seidler (2006) examined younger (18–31 years) and
older participants (65–80 years) in their ability to learn different joystick aiming tasks. Older
adults exhibited poorer precision and took longer time to learn the joystick aiming task as
compared to younger adults. Anshel (1978) compared the learning ability of younger (22–26
years) and older (70–80 years) participants on a limb repositioning task. Younger adults
performed more accurately than older adults. However, with practice, the older participants
showed significantly greater improvement than younger participants. The findings of this
study suggested that the elderly individuals were capable of learning the limb repositioning
task but performance was affected by age-related changes.
Breitenstein, Daum, and Schugens (1996) compared the performance of younger and older
participants on simple and mirror-reversed tracking tasks. In the simple tracking task, the
extent of improvement in performance was similar between the younger and older
participants. But in the more complex mirror reversed tracking task, the performance of the
younger participants was much better than the older counterparts. This suggests that with
increase in complexity of the tasks, the age difference in motor learning is more pronounced.
Similarly, other studies have found that even though older participants tend to learn complex
motor skills, performance accuracy is lower in comparison to younger participants (Wishart,
Lee, Cunningham, & Murdoch, 2002; Wright & Payne, 1985).
The task complexity has been considered a major factor in determining the amount of
motor learning in elderly individuals. Welford (1985) demonstrated that for simple tasks, the
elderly participants learned the task quickly and there was not much difference in terms of
performance accuracy between the younger and older participants. However, Falduto and
Baron (1985) found that on a complex card sorting task, the performance of the younger
participants was much better than the older participants.

55
Speech motor learning
Old age is accompanied by changes like deterioration in the physiological functioning of
oral motor structures (e.g., tongue) (Calhoun, Gibson, Hartley, Minton, & Hokanson, 1992),
breakdowns in speech production (Searl, Gabel, & Fulks, 2002), decreased breath support for
speech (Hoit & Hixon, 1987), atrophy of the vocal folds (Takeda, Thomas, & Ludlow, 2000),
and decrease in articulation rate (Jacewicz, Fox, O’Neil, & Salmons, 2009). The studies
which have compared the performance of younger vs. older adults on speech-motor learning
tasks also suggest that older participants tend to perform poorly in comparison to their
younger counterparts (Sadagopan, 2008; Ballard, Robin, Woodworth, & Zimba, 2001;
Schulz, Stein, & Micallef, 2001). Sadagopan compared the novel speech learning ability in
younger vs. older age groups. A physiological measure (i.e., kinematic analysis) and
behavioural measures (production accuracy and duration) were assessed on two consecutive
days for 16 young and elderly participants during the production of six novel nonwords
increasing in length and complexity. Behaviourally, clear differences were noted between
young and elderly participants in the ability to accurately produce the longer, more complex
nonwords. Older speakers’ productions revealed a significantly greater percentage of
articulatory errors than young adults for four-syllable nonwords, suggesting that important
age-related differences are present for repetition of long, complex novel nonwords. Elderly
individuals also demonstrated longer durations for nonword production than young adults,
and this effect was magnified for longer, more complex nonwords. Very few elderly
individuals were able to produce the requisite number of accurate productions for kinematic
analysis of the two most complex nonwords, and these were excluded from statistical
analyses.
Ballard et al. (2001) investigated the age-related changes in a visuomotor tracking (VMT)
task. In a VMT task related to an articulator, the participant is required to trace the movement
of the target signal using an articulator of interest (like the tongue). In this study, the control
of, lower lip, jaw, and larynx was studied across the life span using the VMT task. A total of
52 females and 35 males ranging in age from 8 to 84 years participated. For the lip and jaw,
then VMT performance was studied using a strain gauge cantilever system. To study the
control of larynx, the participants were required to sustain a vowel (/a/). Results revealed that
the movement accuracy was better in the younger participants in comparison to the older
participants.
Schulz et al. (2001) studied performance of healthy younger and older participants on a
novel speech utterance with respect to the kinematic measurement of the articulators. Three

56
younger males (22-24 years) and three older males (54-68 years) participated in the study. All
the participants practised producing a novel meaningless speech utterance in five blocks of 10
repetitions each. The kinematic movements of the tongue, lower lip, and jaw were measured
during this practice task. The results suggested older, as well as younger, participants
demonstrated learning capabilities of the utterance. However, the younger males were more
accurate in the production of the utterance, made fewer errors, and showed better retention in
comparison to the older participants. The results of this study were in agreement with the
findings of the non-speech-motor learning, suggesting that elderly individuals are capable of
learning motor skills but tend to be less accurate in performance compared to the younger
individuals (Wishart et al. 2002; Wright & Payne, 1985).
Past studies suggest that old age leads to decreases and slower performance of non-speech,
as well as, speech-motor learning tasks. Various brain-based and cognitive models have been
implicated to explain the reduced motor performance of old aged individuals. Past studies
have proven that older individuals are capable of learning novel motor skills but they do it at
a much slower pace than their younger counterparts. A drawback of past studies is that aging
has not been considered in the application of PMLs in speech, as well as, non-speech tasks. It
is possible that systematic application of PMLs in elderly population might facilitate motor
learning to a major extent. It remains to be determined whether the application of PMLs that
facilitates non-speech and speech motor learning in young adults are similar to those for older
individuals.

Spatial and temporal aspects of motor learning

Spatial learning in general, refers to learning the movement characteristics of novel motor
movements in relation to the learner’s surrounding environment and spatial orientation (e.g.,
learning to articulate the sounds of a word correctly, learning to kick a football accurately).
Temporal learning would refer to learning the duration and/or pacing required to perform the
motor skill (Schmidt & Lee, 2005) (e.g., learning to say a phrase within the specified
duration, performing a throwing task at a certain pace). Most studies applying PMLs in
speech and non-speech-motor learning tasks have investigated either the spatial (e.g.,
Shoenfelt at al., 2002; Wulf & Schmidt, 1997; Mass & Farinella, 2012; Hulla et al., 2008;
Knock et al., 2000) or temporal aspects of motor learning (Shea, Lai, Wright, Immink, &
Black, 2001; Sekiya, Magill, & Anderson, 1996; Adams & Page, 2000). However, it is
important to recognize that movements possess both spatial and temporal characteristics.
Even though, spatial and temporal aspects can be dissociated separately (e.g., drawing a

57
square in different sizes and at different speeds), the entirety of motor learning is captured by
measuring both the spatial and temporal aspects of motor learning. To date, there have been
no attempts to apply PMLs to the learning of both spatial and temporal speech or non-speech
movements.
An important concept within the framework of motor learning is the ‘speed-accuracy
trade-off’ (SAT) (Wickelgreen, 1977). In a typical SAT, the speed of the motor skill is
reduced when focus is on accuracy and vice-versa (Schmidt & Lee, 2005). In other words,
movements can be performed very quickly with compromised accuracy, or they can be
performed accurately at the expense of being slower. For example, when a person tries to
insert a key into a keyhole to open a door, he needs to perform the task at a slower pace so
that the accuracy is not compromised. However, many tasks have both speed and accuracy
requirements. An ideal way to approach these tasks is to make movements as fast as possible
without compromising the accuracy. For example, this might be applicable to tasks like
kicking a football or swinging a tennis racket. Both these tasks must be done quickly but also
with precision (Fairbrother, 2010).
The notion of SAT has been implicated in non-speech tasks (Keramati, Dezfouli, & Piray,
2011), as well as in speech tasks, (Goozee, Stephenson, Murdoch, Darnell, & Lapointe, 2005;
Parnell & Amerman, 1996). For example, Goozee et al. compared the lingual kinematics in a
group of younger and older participants. Eight younger females (M = 26.7 years) and eight
older females (M = 67.1 years) were required to repeat /ta/ and /ka/ at a moderate rate and as
fast as possible. Electromagnetic articulography was used to track the lingual movements
during these speech tasks. The results revealed that during the fast speaking condition, both
groups reduced the distance travelled by the tongue. However, older participants
demonstrated a SAT to maintain the accuracy in articulating the sounds. SAT might provide
important information about the approach (spatial vs. temporal) adapted by individuals while
learning a certain motor skill. This information could be useful in teaching novel motor skills
or in designing treatment protocols for various motor disorders. Most studies related to motor
learning have focussed solely on either spatial or temporal learning, thereby not having an
opportunity to observe the SAT situation. Hence, measuring the outcomes of motor learning
in terms of both spatial and temporal aspects would provide a more complete picture about
the process of motor learning.

58
 Statement of the Problem
The notion that PMLs facilitate learning/re-learning of motor skills has largely emerged
from studies involving non-speech-motor tasks (e.g., finger tapping, keyboard entry, arm
stretching and lever positioning tasks) which have been conducted over the past 40 years
(Adams, 1971; Schmidt, 1975b). Identifying the PMLs which tend to be the most effective in
learning non-speech-motor skills or treating motor disorders has been a matter of debate.
Still, the general consensus is that PMLs tend to be effective in learning various facets of
motor skills (e.g., Steinhauer & Grayhack, 2000; Murray & Udermann, 2003; Emanuel,
Jarus, & Bart, 2008). Past studies have validated the usefulness of PMLs in learning non-
speech-motor skills in healthy individuals (Breslin et al., 2012; Rohrer et al., 2005; Murray &
Udermann, 2003; Emanuel, Jarus, & Bart, 2008) as well as in clinical populations (Kwakkel,
2009; Verschueren, Swinnen, & Dom, 1997; Dick et al., 1996; Lin et al., 2007).
Since the last decade, there has been considerable interest to investigate the benefits of
PMLs in relation to speech-motor learning (e.g., Adams & Page, 2000; Knock, et al., 2000;
Steinhauer & Grayhack, 2000; Adams et al., 2002; Ballard et al., 2007; Hula et al., 2008;
Maas et al., 2008; Katz et al., 2010; Wong et al., 2011; Mass & Farinella, 2012; Bislick et al.,
2012). All of the aforementioned studies have been responsible for shifting the focus of
application of PMLs from non-speech tasks to speech tasks. In addition, they have also
provided evidence that PMLs are useful in learning speech-motor tasks, as well as in treating
MSDs. In spite of these advances, there are a number of limitations in past studies that need
to be addressed. Some of the limitations serve as the basis for the present study.
The first major limitation is that past studies have failed to directly compare the effects of
PMLs on both speech and non-speech-motor learning tasks in the same individual. Because
PMLs have been drawn heavily from non-speech (limb-based) tasks, it may be reasonable to
deduce that the PMLs applicable to non-speech-motor learning will also be applicable to
speech-motor learning, as both are motor skills. However, research has shown that limb-
motor control and speech-motor control differ in terms of their physiological nature (Smith,
2006), degrees of freedom, and cognitive requirements (Grimme et al., 2011). Therefore, it is
uncertain whether non-speech motor control and speech-motor control will respond to PMLs
in a similar manner.
A second limitation with past studies of non-speech and speech-motor learning is that they
have failed to consider the combined effects of practice and feedback variables on motor
learning. It is possible that motor learning is effective in situations where an individual

59
receives both practice and feedback simultaneously, so addressing only one of these issues
(i.e., either practice or feedback) in learning a motor skill will limit the motor learning ability
of an individual.
A third limitation is that the past studies have not investigated the effects of PMLs on
aging population. Decreased motor performance is a typical finding in studies on normal
aging (Mattay et al., 2002; Perrot & Bertsch, 2007). It is likely that the PMLs might affect the
elderly individuals in a different manner than the younger age group. Research related to age
effects and PMLs is yet to be undertaken.
The fourth limitation is that past studies have not compared the effects of PMLs in healthy
and in individuals with motor-based disorders. Past studies have revealed that individuals
with motor-based disorders (e.g., Parkinson’s disease) tend to have difficulty in executing
motor activities (Marsden, 1989). It is likely that the application of PMLs in individuals with
motor-based disorders and healthy individuals might influence motor learning to varying
degrees in both the groups. Addressing this issue might help in designing therapy protocols
incorporating PMLs for individuals with motor-based disorders.
Finally, past studies have not addressed motor learning in terms of both spatial and
temporal learning abilities. It is well known that a motor skill is comprised of a spatial and a
temporal domain (Kelso, 1992). Some of the past studies related to temporal learning have
found that participants learn temporal skills at the expense of spatial accuracy (e.g., Goozee
et al., 2005). It is essential to estimate motor learning within the scope of spatial and temporal
domains to fully appreciate the effectiveness of PMLs.
In conclusion, the goals of the current study were: (1) to investigate the effect of selected
principles of motor learning on non-speech and speech-motor learning in individuals with
normal speech-motor control and impaired speech-motor control, and (2) to compare the
spatial and temporal learning of speech, as well as non-speech tasks, in older and younger age
groups.
To achieve these goals the following hypotheses were posed:

Hypothesis 1 – The PMLs that best facilitate spatial learning of a novel musical keyboard
entry task (non-speech task) will also best facilitate spatial learning of a novel speech
utterance (speech task) in a group of healthy individuals.

Hypothesis 2 – The PMLs that best facilitate temporal learning of a novel musical keyboard
entry task (non-speech task) will also best facilitate temporal learning of a novel speech
utterance (speech task) in a group of healthy individuals.

60
Rationale for H1 & H2: Speech is considered to be a dynamic motor system essential for
human communication. The speech and skeletomuscular systems share common neural
control modes despite fundamental biomechanical differences (Perrier, Ostry & Laboissière,
1996). Recent research reveals that that coordination among limb and articulatory effectors
share common physiological framework and have been investigated with similar
experimental methods (Perrier, Ostry & Laboissière, 1996). Hence, the practice and feedback
conditions inducing changes in the non-speech-motor system can be expected to influence the
speech-motor system in a similar manner. Therefore, it is reasonable to hypothesize that
speech production as a motor skill is governed by similar principles of motor learning as the
non-speech-motor system.

Hypothesis 3 – The PMLs that best facilitate spatial learning of a novel speech utterance task
will not be similar between a group of healthy younger individuals and a group of healthy
older individuals.

Hypothesis 4 – The PMLs that best facilitate spatial learning of a novel musical keyboard
entry (non-speech) task will not be similar between a group of healthy younger individuals
and a group of healthy older individuals.

Hypothesis 5 – The PMLs that best facilitate temporal learning of a novel speech utterance
task will not be similar between a group of healthy younger individuals and a group of
healthy older individuals.

Hypothesis 6 – The PMLs that best facilitate temporal learning of a novel musical keyboard
(non-speech) entry task will not be similar between a group of healthy younger individuals
and a group of healthy older individuals.
Rationale for H3-H6: Past studies have revealed differences of the speech and non-speech-
motor systems between younger and elderly individuals (Mattya et al., 2002; Perrot &
Bertsch, 2007). Some common examples of motor performance deficits in older individuals
include difficulty in coordination (Seidler et al., 2002), increased movement variability
(Vidal, Teulings, & Stelmach, 1998) in comparison to younger individuals. Previous research
investigating motor learning in older and younger individuals suggests that the extent of
motor learning tends to vary between older and younger individuals (Fraser, Li & Penhune,
2009). It is possible that the speech and non-speech-motor systems of the younger and elderly
individuals might be influenced to varying extent upon application of PMLs. Although,
investigating the age effect on speech-motor learning seems to be pertinent to the current

61
study, a parallel investigation of the age effect on non-speech-motor learning might provide
valuable evidence pertaining to the aging motor system in general.

Hypothesis 7- The PMLs that best facilitate spatial learning of a novel speech utterance task
will be similar between a group of healthy individuals and a group of individuals with
hypokinetic dysarthria due to PD.

Hypothesis 8 - The PMLs that best facilitate spatial learning of a novel musical keyboard
(non-speech) entry task will be similar between a group of healthy individuals and a group of
individuals with hypokinetic dysarthria due to PD.

Hypothesis 9 - The PMLs that best facilitate temporal learning of a novel speech utterance
task will be similar between a group of healthy individuals and a group of individuals with
hypokinetic dysarthria due to PD.

Hypothesis 10 - The PMLs that best facilitate temporal learning of a novel musical keyboard
entry (non-speech) task will be similar between a group of healthy individuals and a group of
individuals with hypokinetic dysarthria due to PD.
Rationale for H7-H10: There is evidence to show that individuals with PD are also capable
of motor learning (Soliveri, Brown, Jahanshahi, & Marsden, 1992; Behrman, Cauraugh, &
Light, 2000). It is likely that the PMLs which influence the speech and non-speech-motor
systems of healthy individuals will also influence the speech and non-speech-motor systems
of individuals with PD to a similar extent. However, the performance of individuals with PD
can be expected to be reduced in comparison to the healthy counterparts.

62
 Chapter 3. Methods

Participants
Non-clinical group - The study involved two experiments. The first experiment involved
recruitment of a non-clinical group of 80 healthy individuals (21 males & 59 females) in the
age range of 40-80 years (M = 59 years). The inclusion criteria for the participants were (1)
no reported history of sensory and cognitive abnormalities, (2) native speaker of New
Zealand English, (3) completion of a high school diploma, and (4) right-hand dominance.
The right hand dominance was assessed using the Edinburgh Handedness Inventory
(Oldfield, 1971) prior to the start of the data collection. The inventory provides a quantitative
index of handedness. Details of the inventory are provided in Appendix 3. The participants
for the non-clinical group were recruited through a convenience sampling procedure from the
database of control participants registered with the New Zealand Brain Research Institute
(Christchurch, NZ) and also from the wider community of Christchurch. The demographic
details of the participants in the non-clinical group are presented in Table 4.

Clinical group - The second experiment involved recruiting a clinical group of 16

individuals (12 males & 4 female) in the age range of 58-84 years (M = 70 years) with
hypokinetic dysarthria due to PD. The participants were recruited from a list of patients
receiving support services from the Parkinson’s Society of New Zealand (Canterbury
Division). Based on the examination of medical records, the condition of PD was diagnosed
as a chronic progressive syndrome with two of three cardinal features of rest tremor,
bradykinesia, and rigidity, without evidence of a secondary cause or atypical features. The
onset of the PD for the participants ranged from 4-12 years. All participants were on
dopamine replacement therapy. The data collected in the present study occurred while the
participants were in a self-reported ‘on’ state, 1 to 4 hours after taking medications. The ‘on’
state refers to the time period following medication which represents the best motor ability of
patients with PD (Goberman & Coelho, 2002). The Movement Disorder Society–Unified
Parkinson’s Disease Rating Scale (MDS-UPDRS) ‘motor examination’ subsection scores
(Goetz., et al, 2007), and modified Hoehn and Yahr staging (Goetz et al., 2004) were
recorded for each participant with PD on the first day of the experiment. The ‘motor
examination’ subsection of the MDS-UPDRS consists of 18 evaluation parameters.

63
Table 4. Descriptive data of the participants in the non-clinical group including age, sex, and
practice conditions. Mean age of the participants is indicated at the bottom of the table.

Participants Age Sex Practice condition

1 78 F Constant
2 73 F Constant
3 71 F Constant
4 71 F Constant
5 71 F Constant
6 67 F Constant
7 64 F Constant
8 64 F Constant
9 62 F Constant
10 61 F Constant
11 60 M Constant
12 57 F Constant
13 57 F Constant
14 56 F Constant
15 56 F Constant
16 53 F Constant
17 51 F Constant
18 48 F Constant
19 46 F Constant
20 46 M Constant

21 72 F Variable
22 71 F Variable
23 70 M Variable
24 65 M Variable
25 64 M Variable
26 63 F Variable
27 62 M Variable
28 61 F Variable
29 60 F Variable
30 59 F Variable
31 57 F Variable
32 56 F Variable
33 55 F Variable
34 55 F Variable
35 54 M Variable
36 52 F Variable
37 51 F Variable
38 49 F Variable
39 44 F Variable
40 42 F Variable

64
 65

Participants Age Sex Practice condition

41 73 F Random
42 73 M Random
43 72 M Random
44 71 F Random
45 70 M Random
46 65 F Random
47 65 M Random
48 63 F Random
49 62 F Random
50 62 F Random
51 57 M Random
52 55 F Random
53 55 F Random
54 52 F Random
55 52 F Random
56 52 M Random
57 50 F Random
58 49 M Random
59 46 F Random
60 44 M Random

61 75 F Blocked
62 72 F Blocked
63 72 F Blocked
64 71 M Blocked
65 64 F Blocked
66 64 M Blocked
67 64 M Blocked
68 63 F Blocked
69 62 F Blocked
70 61 M Blocked
71 59 M Blocked
72 58 F Blocked
73 58 F Blocked
74 57 F Blocked
75 57 F Blocked
76 57 F Blocked
77 52 F Blocked
78 48 F Blocked
79 45 F Blocked
80 44 M Blocked

M 59.5

65
Each parameter is evaluated by the clinician and given a score ranging from ‘0’ (normal) to
‘4’ severe) based on the performance of the participant on a particular task. The total score
for the ‘motor examination’ subsection ranged from ‘0’ (normal) to ‘82’ (severe). The
average UPDRS score during the ‘on’ state was 45.3 (range = 21-61, SD = 12.2) indicating
moderate motor impairment, and the average speech UPDRS subscale rating was 1.56 (range
= 1–3, SD = 0.62) indicating mild-moderate speech impairment. The modified Hoehn and
Yahr staging scale evaluates the severity of PD based on five separate stages, with 0.5
increments between each stage. Stage one indicates unilateral signs and symptoms and stage
five indicates severe impairment requiring total assistance. The average Hoehn and Yahr
stage during the ‘on’ state testing was 1.8, indicating the stage of bilateral involvement. In
addition, all the participants were administered the Frenchay Dysarthria Assessment (FDA)
(Enderby, 1983) to arrive at a diagnosis of dysarthria. The FDA provides a standardized
assessment of speech neuromuscular activity, involving respiration, phonation, resonance and
articulation, and speech-related reflex activity. The FDA allows the clinician to rate the
ability of each speech subsystem using a 9-point scale, and thus provides a profile that
contributes to the differential diagnosis of dysarthria. The results of FDA indicated that all
the participants exhibited hypokinetic dysarthria, which is usually associated with PD (Duffy,
2005). Demographic details, UPDRS scores and modified Hoehn and Yahr staging scores of
the participants collected prior to the start of the experiment are presented in Table 5.

Procedure
Four practice conditions and a combination of feedback conditions were applied to the
speech and non-speech-motor learning tasks. The same conditions and tasks were evaluated
in both experiments. The four practice conditions were (1) constant practice, (2) variable
practice, (3) blocked practice, and (4) random practice. The feedback which was provided
constituted a combination of low-frequency, KR, KP, and delayed feedback conditions. The
four practice conditions were paired with feedback conditions in both experiments.

Experiment I: Non-clinical group

Speech Task
Four different groups were organized on the basis of the practice condition (constant,
variable, blocked, & random). Each of the 80 participants was randomly and equally assigned
to one of the four practice conditions, thus constituting 20 participants in each practice
condition (see Table 4). The participants in each practice condition were trained

66
Table 5. Descriptive data of the participants in the clinical group including age, sex, UPDRS scores, Hoehn & Yahr staging scores, and practice
conditions. Mean scores are indicated at the bottom of the table.

Participants Age Sex Motor UPDRS Speech UPDRS Hoehn & Yahr Practice conditions

1 84 M 48 2 2 Constant
2 80 M 53 1 1.5 Constant
3 69 M 60 3 2 Constant
4 71 F 61 2 2 Constant

5 74 M 41 1 1.5 Variable
6 62 M 59 2 3 Variable
7 71 M 34 2 1.5 Variable
8 57 F 32 1 1 Variable

9 71 M 32 2 1.5 Random
10 67 M 42 2 1.5 Random
11 71 M 39 1 2 Random
12 69 M 40 1 1.5 Random

13 71 F 21 1 1 Blocked
14 58 M 61 2 2.5 Blocked
15 81 M 55 1 2.5 Blocked
16 64 M 48 1 2 Blocked

M 70 45.3 1.57 1.8

67
speech phrase “Thak glers wur vasing veen arad moovly”. A meaningless phrase was chosen
as the stimulus for training based on the Challenge Point Framework (CPF) (Guadagnoli &
Lee, 2004).
According to CPF, a more challenging environment facilitates better motor learning in
comparison to a less challenging environment. To make the speech task challenging, a speech
phrase containing a string of non-words which followed the phonotactic rules of the English
language was created. Moreover, learning a meaningless material has shown to be more
difficult in comparison to learning a meaningful material (Epstein, 1962). In addition, the
temporal spacing between the non-words was also altered in an attempt to make the phrase
more challenging. In total, the speech phrase used for training (also called the target phrase)
consisted of 7 non-words and 28 phonemes, in total. Among the 7 non-words, 3 were bi-
syllabic and the rest were monosyllabic. The whole phrase was split into three segments
based on temporal pauses which served as the boundaries. The first segment consisted of two
non-words (of eight phonemes), the second segment consisted of three non-words (of 11
phonemes) and the third segment consisted of two non-words (of nine phonemes). Between
the first two segments there was a pause of 4 seconds (s) and between the second and third
segments there was a pause of 2 s. The temporal durations of the first, second and third
segments were 1 s, 3.1 s and 1.6 s, respectively. The overall duration of the target phrase was
11.7 s. The target phrase used for the purpose of training across all four practice groups is
depicted in Figure 2. The study took place in a sound-treated laboratory in the Department of
Communication Disorders. The procedures used for practising the speech task were similar
across the four practice conditions, although the nature of practise varied. Participants were
seated comfortably in a chair in front of a 19 inch computer screen.
Participants in each of the four practice groups were involved in a practice regime of 50
trials per day, for a two-day period to learn the target phrase.1 The practice trials on each day
began after a pre-practice session. During the pre-practice session, the participants were well
motivated, and were provided with clear instructions on specific goals to be achieved during
the practice regime. During the practice regime, the production of each trial by the
participants was always preceded by the provision of orthographic and auditory
representations of the target phrase. The orthographic representation of the target phrase was
displayed on the computer screen along with the auditory representation. The auditory
representation of the phrase was provided through loud-speakers using a pre-recorded adult
male voice. To assist the participant in the practice regime, the speech phrase was presented
1
An initial pilot study was conducted by engaging the participants in a practice regime of 100 trials each
day during the two-day period. It was found that the participants found it difficult to practise 100 trials
68
each day due to fatigue. Hence, it was determined that 50 practice trials were adequate to engage the
participants in a practice regime.
Figure 2. Target phrase used for training the participants in each practice condition. The target phrase consisted
of 3 bi-syllabic and 4 monosyllabic non-words. In total, there were 28 phonemes. The temporal components of
the phrase are shown.

69
via a Power Point format along with the orthographic and auditory representations displayed
on the computer monitor. The initiation of the Power Point presentation enabled the
participants to see and hear the orthographic and auditory representation of the speech phrase,
respectively. The complete production following the provision of orthographic and auditory
representations comprised one practice trial. Once the researcher judged a trial to be
performed, he pressed the ‘return’ key on the computer keyboard which resulted in the
initiation of the next practice trial. This was carried out until the completion of 50 trials on
each day during the two-day period. Instances of false start allowed the participants to re-start
the production of speech phrase during any particular practice trial. The general equipment
layout for the speech task is shown in Figure 3.
The participants were instructed to match their production to the target phrase as
accurately as possible. At the conclusion of every 10th trial, there was a break of approximate
five minutes during which the researcher performed an acoustic analysis of the participant’s
production of the 10th trial and provided feedback on their performance and accuracy of
production. The researcher measured the overall duration of the phrase, as well as the
individual duration of each segment and pause duration between the segments. In addition to
this temporal analysis, the researcher also perceptually assessed the articulatory accuracy of
the phonemes produced by the participants. The feedback was provided to the participants by
displaying the target phrase on a sheet of A4 paper as shown in Figure 2. The researcher
indicated whether the participant’s production matched the target phrase in terms of temporal
and articulatory accuracy. When providing verbal feedback to the participant on each
temporal feature (overall duration, segment duration and pause duration), the researcher used
terms like “accurate”, “too long” and “too short” in reference to the orthographic rendition of
the phrase. In terms of feedback on articulatory accuracy, the researcher perceptually
analysed the participants’ productions and indicated whether the individual phonemes were
articulated correctly in comparison to the target phrase. Thus, the nature of feedback
provision was low-frequency and delayed, and also included information about KR and KP.
Participants were also informed if there were any instances of misarticulation. Across
the two days there were a total of 100 trials, and each participant received 10 feedback trials
of the target phrase. The participant utterances after each practice trial were audio-recorded
using a desktop condenser microphone (DSE-PC). The output acoustic signal from the
microphone was fed into a laptop computer (Lenovo ideapad S10e) running Audacity 1.3
(Beta version) acoustic analysis software program. The entire experiment took place over
three consecutive days. The first two days served as the acquisition phase of speech-motor

70
Figure 3. Experimental set up for the speech learning task. The computer monitor and the headphones are used
for the purpose of visual and auditory representations of the speech phrase, respectively.

71
learning, during which the participants practised 50 trials of the target phrase each day. The
acquisition phase lasted for 60-90 min on each of the first two days. On the third day, the
participants returned to the laboratory and were seated in front of the computer screen.
However, the participants were not provided with any sort of visual or auditory
representations of the target phase and were required to produce 10 trials of the target phrase
without any further practice or feedback. This formed the retention phase and lasted for 10-15
min. These trials were recorded and stored for later acoustic analyses.

Practice Design for constant practice group - Constant practice was defined as practice on
only one variant of a movement (Mass et al., 2008). During the acquisition phase, the
participants practised 50 trials of the target phrase in a repeated manner each day. After each
ten trials, the participants were provided with feedback as described previously. This resulted
in provision of five feedback trials to participants on each day of the acquisition phase.
Participants returned to the lab on the third day for the retention phase. The participants were
required to recall and produce the previously practised target phrase. The participants
produced 10 trials of the target phrase and no feedback was provided during or after the
retention trials. These trials were recorded for later acoustic analyses.

Practice Design for variable practice group - Variable practice was defined as practice on
more than one variant of a given movement (e.g., practising a golf swing over varying
distances from the hole). Variable practice consists of practising motor tasks which share the
same motor plan but differ in parameter (Mass et al., 2008). In an attempt to include tasks of
similar motor plan but of different parameters, an alternate phrase of different temporal
duration was chosen. Participants practised 25 trials of the target phrase and 25 trials of an
alternate phrase (50 trials in total) during each day of the acquisition phase. The alternate
phrase used the same sequence of non-words; however, the phrase was modified temporally
by changing the duration of the non-words and inter-segment pauses. The phrase was
temporally modified by inserting a pause of 3 s between the first two segments and a pause of
5 s between the second and third segments. The durations of the first, second and third
segments were 2 s, 2 s and 1.6 s, respectively. The overall duration of the target phrase was
13.6 s. The target and alternate phrase used by the participants undergoing variable practice is
shown in Figure 4. The target and alternate phrases were randomized across 100 trials.
Feedback was provided after every 10 trials and the procedure of feedback provision was
similar to constant practice. At the conclusion of 50 practice trials on the first day, there were
three instances when the 10th trial was the target phrase and another two instances when the

72
 (A)

(B)

Figure 4. The target and alternate phrases used for the variable practice condition. The top panel (A) depicts the
target phrase and the bottom panel (B) depicts the alternate phrase. The temporal components of both the
phrases are also shown.

73
10th trial was the alternate phrase. Similarly, at the end of the 50 trials on the second day,
there were three instances when the 10th trial was the alternate phrase and another two
instances when the 10th trial was the target phrase. This arrangement resulted in each
participant receiving five feedback trials of the target phrase and another five feedback trials
of the alternate phrase during the acquisition phase. The order of the feedback trials for the
target and alternate phrases was randomized and counterbalanced across both days of the
acquisition phase and also across the participants to avoid order effect. Participants returned
to the lab on the third day for the retention phase. Participants were required to recall and
produce the previously practised target and alternate phrases. Participants produced 10 trials
of the target phrase and 10 trials of the alternate phrase. Feedback was not provided during or
after the retention trials. These trials were recorded and stored for later acoustic analyses. The
recall order of the retention trials for the target and alternate phrases was counterbalanced
across the 20 participants.

Practice Design for random practice group - Random practice was defined as a practice
schedule in which different movements are produced on successive trials, and where the
target for the upcoming trial is not predictable to the learner (Mass et al., 2008). All the
participants assigned to this group practised 25 trials of the target phrase and 25 trials of a
second alternate phrase during each day of the acquisition phase. The alternate phrase used
for random practice was called the ‘second alternate phrase’ to distinguish this phrase from
the alternate phrase used in the variable practice condition. The second alternate phrase was
“Ang haky deebs reciled tofently roovly”. The nature of random practice involves practising
tasks of a different motor plan. In an attempt to change the motor plan, the second alternate
phrase differed from the target phrase in terms of phonemic composition and temporal
duration. The second alternate phrase consisted of six non-words and 29 phonemes, in total;
of which two were monosyllabic, three were bisyllabic and one was trisyllabic. The whole
phrase was split into three segments based on temporal pauses which served as the
boundaries.

The first segment consisted of two non-words (of seven phonemes), the second segment
consisted of three non-words (of 17 phonemes), and the third segment consisted of one non-
word (of five phonemes). Between the first two segments there was a pause of 2 s and
between the second and third segments there was a pause of 3 s. The temporal duration of the
first, second and third segments were 1.95 s, 2.1 s and 0.57 s, respectively. The overall
duration of the target phrase was 9.62 s. The target and second alternate phrases used for

74
training participants is shown in Figure 5. An equal number of target and second alternate
phrases were randomized across 100 trials. The manner of feedback provision was similar to
the other practice conditions. The organization of the feedback trials was similar to the
variable practice condition. At the end of the acquisition phase, there were five instances
when the 10th trial was the target phrase and another five instances when the 10th trial was the
second alternate phrase. This arrangement resulted in each participant receiving five feedback
trials of the target phrase and another five feedback trials of the alternate phrase during the
acquisition phase. The order of the feedback trials was randomized and counterbalanced
across both days of the acquisition phase and also across the participants. The retention phase
on the third day also followed a similar procedure as the variable practice condition. The
recall order of the retention trials for the target and second alternate phrases was
counterbalanced across all the 20 participants.

Practice Design for blocked practice group – Blocked practice was defined as a practice
schedule in which the learner practises a group of the same target movements before
beginning practice on the next target (Mass et al., 2008). The participants practised the target
phrase and the second alternate phrase in blocks of 25 trials each. This resulted in participants
practising the target phrase from trials 1 through 25, and the second alternate phrase was
practised from trials 26 through 50, and vice-versa on the second day. Due to this
arrangement of blocked practice, the feedback at the end of the 30th trial was based on the
phrase which was practised from trials 25 through 30. For example, if the target phrase was
practised from trials 26 through 30, then the feedback for the target phrase was provided at
the end of the 30th trial, and this was reversed on the second day. The procedure of feedback
provision was similar to the other practice conditions. This resulted in each participant
receiving five feedback trials of the target phrase and five feedback trials of the second
alternate phrase at the end of the acquisition phase. The order of the feedback trials for the
target and second alternate phrases was randomized and counterbalanced across both days of
the acquisition phase and also across the participants to avoid order effect. The retention
phase on the third day followed a similar procedure as the random practice condition.

75
 (A)

(B)

Figure 5. The target and second alternate phrases used for the random practice condition are shown. The top
panel (A) depicts the target phrase and the bottom panel (B) depicts the second alternate phrase. The temporal
components of both the phrases are also shown.

76
 NASA Task Load Index - Following the retention phase on the third day, each of the 80
participants were required to complete the NASA Task Load Index (Hart & Staveland, 1988)
to assess the complexity of the speech task. The NASA Task Load Index is a multi-
dimensional rating scale used to assess the overall workload associated with a given
performance situation. The index evaluates the workload in terms of six different subscales:
mental demand, physical demand, temporal demand, own performance, effort, and
frustration. Each subscale consists of 21 gradations with one extreme representing ‘very low
demand’ and the other extreme representing ‘very high demand’. The index provides an
overall workload score based on an average of these six subscales. The participants were
instructed to indicate their preference across each subscale by putting a check across the
appropriate gradation. The ratings across these six subscales were averaged to obtain a mean
perceived difficulty rating score. The index is shown in Figure 6.

77
Figure 6. NASA Task Load Index (Hart & Staveland, 1988) which assesses workload on six 7-point scales.
Increments of high, medium and low estimates for each point result in 21 gradations on the scales.

78
Non-speech Task
Each participant also completed a non-speech task and was assigned to the same practice
groups as the speech task. The non-speech task involved training on a musical keyboard. The
task required the participants to use the index finger of the (dominant) right hand to practise a
musical tune (also called the target tune) comprised of a sequence of musical notes. The non-
speech task occurred on the same days and session as the speech task. The sequence of
practising the speech and non-speech tasks was counterbalanced across the participants to
avoid any order effect. The target tune was “FBG#A# FG#AA# FG#A#B”. Similar to the target
(speech) phrase, the entire target tune was organized into three musical segments. Each
segment consisted of three musical notes, with a total of 12 notes across the tune. The
duration of the first, second and third segments were 1.75 s, 1.7 s and 1.8 s, respectively.
There was a pause of 4 s between the first and second segments and a pause of 2 s between
the second and third segments. The entire duration of the target tune was 11.29 s. For the
purpose of visual representation, the musical notes in each segment were depicted by dots of
increasing size on the keys. The size of the dots indicated the order of keys to be pressed on
the keyboard (e.g., the smallest dot on the key would indicate the first key to be pressed, and
so on). The target musical tune is illustrated in Figure 7.
Similar to the speech task, participants in each of the four practice conditions were
required to undergo a practice regime of 50 trials during each day of the acquisition phase to
learn the target tune. The practice trials on each day began after a pre-practice session.
During the pre-practice session, the participants were instructed about the nature of the task
and the expected outcomes during the practice regime. The non-speech task followed the
same procedure as the speech task. The production of each trial by the participants was
always preceded by the provision of visual and auditory representations of the target tune.
The visual representation of the target tune was displayed on a computer monitor along with
an auditory representation of the tune. A pre-recorded target musical tune delivered through
loud-speakers served as the auditory representation.
The 50 practice trials during each day of the practice regime were visually presented to
each participant on a Power Point format along with an auditory representation of the target
tune. The complete production of the keyboard tune after the provision of visual and auditory
representations comprised one practice trial. Once the researcher judged a trial to be
performed, he pressed the ‘return’ key on the computer keyboard which resulted in the
initiation of the next practice trial. This was carried out until the completion of 50 trials on
each day during the two-day period. If a participant was observed to exhibit a false start,

79
Figure 7. Target musical tune used for training the participants in each practice condition. The size of the dots
indicates the playing order of the musical notes in each segment. The smallest dot on the key indicates the first
key to be pressed, and so on. The temporal components of the tune are also shown.

80
he/she was allowed to re-play the musical tune during any particular practice trial. Keyboard
entry productions after every practice trial were audio-recorded in a manner similar to the
speech task. At the conclusion of every 10 practice trials, the participants received an
approximate five minutes break during which they received feedback which was low-
frequency and delayed in nature, and also included information about KR and KP. The
feedback on temporal accuracy was similar to the speech task. Feedback on production
accuracy involved the researcher perceptually analyzing the keyboard entry productions and
informing the participants of any incorrect keyboard entry productions. The acquisition phase
lasted for 60-90 min on each of the first two days and the retention phase lasted for 10-15
min. The retention phase was identical to the speech task. The experimental setup for the
keyboard task is shown in Figure 8.

Practice design for constant practice group - During each day of the acquisition phase, the
participants practised 50 trials of the target tune in a repeated manner. The nature and
provision of feedback was provided in a manner similar to the constant practice condition in
the speech task. Participants returned to the lab for the retention phase and were required to
recall and reproduce 10 trials of the target tune without further practice or feedback. These
trials were recorded for later acoustic analyses.

Practice design for variable practice group - The participants practised 25 trials of the
target tune and 25 trials of an alternate tune during each day of the acquisition phase. An
alternate tune of different duration was chosen in an attempt to practise tasks of similar motor
plan but of different parameters. The alternate tune used for variable practice consisted of the
same number and sequence of musical notes in each segment as the target tune, but differed
in duration. The durations of the first, second and third segments were 1.75 s, 1.7 s and
1.84 s, respectively. There was a pause of 3 s between the first and second segments and a
pause of 5 s between the second and third segments. The entire duration of the target tune
was 13.29 s. The target and alternate tunes were randomized across 100 trials. The target and
alternate tunes used for training participants is shown in Figure 9. Randomization of the
feedback trials and the procedure for feedback provision was similar to the variable practice
condition in the speech task. During the retention phase, participants produced 10 trials of the
target tune followed by another 10 trials of the alternate tune. These trials were recorded and
stored for later acoustic analyses. The recall order of the retention trials for the target and
alternate tunes were counterbalanced across the 20 participants.

81
Figure 8. Experimental set-up for the keyboard task. The computer monitor and the headphones are used for the
purpose of visual and auditory representations of the musical tune, respectively.

82
(A)

(B)

Figure 9. The target and alternate tunes used for the variable practice condition. The top panel (A) depicts the
target tune and the bottom panel (B) depicts the alternate tune. Both tunes consist of same sequence of four
musical notes in each segment. The size of the dots indicates the playing order of the musical notes in each
segment. The smallest dot on the key indicates the first key to be pressed, and so on. The temporal components
of both the tunes are also shown.

83
 Practice Design for random practice group - Participants practised 25 trials of the target
tune and 25 trials of a second alternate tune during each day of the acquisition phase. To
distinguish the alternate tune used in the variable practice condition, the alternate tune in the
random practice condition was called the ‘second alternate tune’. The second alternate tune
was “FA#G#G A#G#FF # GAFA#”. The second alternate tune differed from the target tune in
terms of musical notes and temporal duration. The alternate tune consisted of three segments
separated by pauses. Each segment in turn consisted of a sequence of four musical notes.
Between the first two segments there was a pause of 2 s and between the second and third
segments there was a pause of 3 s. The durations of the first, second and third segments were
1.87 s, 2.17 s and 2.05 s, respectively. The overall duration of the second alternate tune was
11.09 s. The target and second alternate tunes used for training participants is shown in
Figure 10. An equal number of target and alternate phrases were randomized across 100
trials. Feedback was provided in a manner similar to the random practice condition in the
speech task. The retention phase on the third day also followed the same procedure as the
speech task.

Practice design for blocked practice group - Participants practised the target tune in a
block of 25 trials followed by a block of 25 trials comprising the second alternate tune (used
for the random practice condition) or vice versa during each day of the each day of the
acquisition phase. The order of blocked practice was counterbalanced across 20 participants
to avoid an order effect. The feedback provision was of the same nature as the blocked
practice condition in the speech task. The retention phase on the third day also followed a
similar procedure as the blocked practice condition in the speech task.

NASA Task Load Index - Similar to the speech task, each participant was required to
complete the NASA task load index following the retention phase to assess the complexity of
the non-speech task.

84
(A)

(B)

Figure 10. The target and second alternate tunes used for the random practice condition. The top panel (A)
depicts the target tune and the bottom panel (B) depicts the second alternate tune. The second alternate tune
differs from the target tune in terms of the musical notes and temporal duration. The size of the dots indicates
the playing order of the musical notes in each segment. The smallest dot on the key indicates the first key to be
pressed, and so on. The temporal components of both the tunes are also shown.

85
Experiment II: Clinical group
Experiment II involved a clinical group of 16 participants with PD. The experiment
procedures were identical to Experiment I. The 16 participants were randomly and equally
assigned to the four practice conditions, resulting in four participants in each practice group.
The speech and non-speech tasks were carried out in the same fashion as Experiment I. The
speech and non-speech tasks occurred on the same days and the order of the learning tasks
was counterbalanced across the 16 participants to avoid any order effect. The experiment
took place over three consecutive days. The first two days of the experiment served as the
acquisition phase and the third day was the retention phase.

Data Analyses
The final five trials of the 10 trials obtained from each participant for the speech and non-
speech tasks during the retention phase were analysed. The reason for including only the final
five retention trials for data analysis was that of all the ten trials produced by the participants
during the retention phase, the last five trials represented better production accuracy in
comparison to the first five trials as judged by the researcher. Also, this method of including
the final five trials of the participant productions for data analysis has been previously
reported in a study (Adams & Page, 2000). The data were analysed according to (1) spatial
and (2) temporal features of production accuracy. The data analysis procedures were similar
for clinical and non-clinical groups.

Spatial analysis
Speech task - Spatial analysis of the speech task involved evaluating the production
accuracy of the target speech phrase by calculating the Percentage of Phonemes Correct
(PPC) (Dollaghan & Campbell, 1998). The production accuracy of the alternate and second
alternate phrases was not evaluated as the focus of the study was on the learning outcome of
the target phrase alone. The PPC is calculated by dividing the number of correct phonemes
produced by the total number of phonemes produced and multiplying by 100. The mean PPC
obtained from the final five retention trials was calculated for each participant. The individual
mean values were then averaged across the 20 participants in each practice group to obtain a
grand mean PPC.
Keyboard task - Spatial analysis of the keyboard learning task was based on the
calculation of Percentage of Keystrokes Correct (PKC). The PKC was devised in a manner
similar to PPC to evaluate the production accuracy of the target tune during the retention
trials. The PKC is calculated by dividing the number of correct keystrokes produced by the

86
total number of keystrokes produced and multiplying by 100. The final five retention trials
for each participant were used in the calculation of a mean PKC for each participant. The
individual mean values were then averaged across the 20 participants in each practice group
to obtain a grand mean PPC.

Temporal analysis
Temporal analysis focused on calculating the temporal synchrony of the participant
productions during the retention trials of the speech phrase and keyboard tune in comparison
to the original examples of the target phrase and tune, respectively. The final five trials of the
target phrase/tune produced by each participant during the retention phase were submitted to
acoustic analysis. The acoustic analysis was carried out using Audacity 1.3 (Beta version).
The participant productions during the retention trials, as well as the original examples of the
target phrase and tune were digitized at a 44 kHz sampling rate and simultaneously displayed
one below the other on a computer monitor as amplitude-by-time waveforms. To determine
the synchrony between the participants’ production of the target phrase/tune in comparison to
the original examples of the target phrase/tune, the participants’ productions of the target
phrase/tune were acoustically aligned to the original phrase/tune. The alignment of the
participant productions and the original example of the target phrase/tune occurred at the
onset point of the acoustic waveform. The participant productions and the original target
waveform shared the same onset point. The offset point was based on offset of the original
target waveform. So if the participant’s production was longer than the target waveform, then
the part of the waveform that exceeded the offset point was excluded from the analysis. Once
the waveform of participants’ production and the original target waveform were aligned
according to the onset and offset points of the original target phrase/tune, a pair of vertical
cursors was placed at the onset and offset points. The part of the two waveforms between the
vertical cursors was converted to binary values. The process of converting the waveforms to
binary plots was carried out through a Matlab based program. The steps involved in
converting the acoustic waveforms to binary variables were as follows:

a. The waveforms of the target phrase and tune as well as the waveforms of the
participant productions during the retention trials were digitized at a sampling rate of
44 KHz. This yielded 514800 samples and 496760 samples for the target phrase and
tune, respectively. The number of samples for the participant productions ranged from
264,000 to 500,000.

87
b. The next step involved rectifying and smoothing the target waveforms and that of the
waveform of the participant productions.
c. Following the rectification and smoothing, a threshold was set at 10% of the whole
waveform’s amplitude (i.e. for the entire utterance/keyboard tune) to arrive at the
binary values. The portion of the waveform that was above the threshold was
converted to 1’s and part of the waveform below the threshold value was converted to
0’s. The amplitude of the extraneous noises (like heavy breathing) was revealed to
exceed the 10% of the waveform’s amplitude, so setting a cut-off threshold of 10%
limited the inclusion of the extraneous noises in the signal.
d. The binary values yielded a plot for the original target phrase/tune and participant’s
production. These binary values were used to calculate the phi correlation between the
participant productions and the target phrase/tune. The steps (a-d) involved in
converting an acoustic waveform of a keyboard entry production to a binary plot is
depicted in Figure 11.
A phi correlation was used to assess temporal relation (synchrony) between the two
signals. The phi correlation is a measure of the degree of association between two binary
variables (Field, 2010). A phi correlation was obtained from each of the final five
responses and these values were averaged to obtain a mean phi correlation for each
participant. A grand mean phi correlation value was calculated for the 20 participants in
each of the four practice conditions. The temporal synchrony between the waveform of a
keyboard production during the retention trial and the waveform of the target tune is
illustrated in Figure 12.

88
(A)

(B)

(C)

Figure 11. Conversion of acoustic waveform to binary variables for calculation of phi correlation. The
top panel (A) depicts the raw acoustic waveform of a keyboard entry production. The middle panel (B)
depicts the waveform after being subjected to rectification, low-pass filtering, and smoothing. The
horizontal black line denotes the threshold set at 10% of the waveform’s amplitude. The bottom panel
(C) shows the binary plot used to calculate the phi correlation.

89
(A)

(B)

(C)

(D)

Figure 12. An illustration of temporal synchrony between the waveform of a keyboard production during the
retention trial and the waveform of the target tune. Panel (A) depicts the waveform of the target musical tune
after rectification and low-pass filtering, the red trace indicates the smoothed version of the waveform, and the
black horizontal line depicts the 10% threshold. Panel (B) depicts the waveform of the keyboard entry
production after rectification and low-pass filtering, the red trace indicates the smoothed version of the
waveform, and the black horizontal line depicts the 10% threshold. Panel (C) depicts the binary plots of the
waveforms of the target tune and the keyboard entry production during the retention trial. Panel (D) depicts the
binary plots of the temporal match between the two waveforms. This binary plot yielded a phi correlation of
0.48.

90
Statistical analyses
Four different analyses were carried out. The first set of analysis compared the spatial and
temporal learning among the practice conditions for speech and keyboard tasks. For the
spatial analysis, the mean PPC and PKC scores obtained during the retention phase trials
across all the participants were subjected to a series of two-way mixed model ANOVAs (2
practice tasks X 4 practice conditions). The between group factor was the practice condition
(constant, variable, random, & blocked practice conditions) and the within-group factor was
the practice task (speech & keyboard tasks). Similarly for the temporal analysis, the mean phi
correlation values for the speech and keyboard learning tasks obtained during the retention
phase trials across the participant groups were subjected to a series of two-way mixed model
ANOVAs.
The second analysis compared the effect of older and younger age groups on speech and
keyboard tasks with regard to spatial and temporal learning. A median split was performed to
split the participants into younger and older age groups. The spatial analysis involved
subjecting the mean PPC/PKC of the older and younger age groups to a two-way ANOVA (2
age groups X 4 practice conditions). Similarly, the temporal analysis involved subjecting the
mean phi correlation values of the younger and older age groups to a two-way ANOVA (2
age groups X 4 practice conditions).
The third analysis compared the clinical and non-clinical groups on speech and
keyboard learning tasks by subjecting the PPC/PKC and phi correlation values to a two-way
ANOVA (2 groups X 4 practice conditions). The fourth and final analysis compared the
perceived difficulty among the practice conditions in speech and keyboard learning tasks by
subjecting the mean NASA task load index scores of the younger and older age groups to a
two way ANOVA. In addition, to compare the perceived difficulty of speech vs. keyboard
learning tasks, the index scores across all the four practice conditions were collapsed and
were compared using a paired sample t- test.
The ANOVA tests yielded a F value, p value, and a partial eta squared value to calculate
the effect size. Partial eta squared is the ratio of variance accounted by an effect and that
effect plus its associated error variance within an ANOVA study (Brown, 2008). The
guidelines recommended by Barnette (2006) were used to relate the partial eta squared values
to effect size. One percent of the variance accounted by the predictor variable relates to small
effect size (0.2), a variance of 6% accounted by the predictor variable relates to medium
effect size (0.5), and a variance of 14% accounted by the predictor variable relates to large
effect size (0.8).

91
Measurement reliability
Intra-rater reliability was judged for the spatial and temporal analyses. Specifically, for the
spatial analysis, the mean PPC and PKC values during the retention phase were re-analysed,
and for the temporal analysis, mean phi correlation values during the retention phase were re-
analysed. The measurement reliability was performed for the non-clinical as well as clinical
groups by randomly choosing 20% of the data (i.e., 16 of 80 participants in the non-clinical
group, and 4 of 16 participants in the clinical group).
The Pearson Product Moment Correlation was used to calculate the intra-judge reliability.
For the non-clinical and clinical groups, the intra-rater reliability of the spatial analysis
ranged from r = 0.91 to r = 0.99. The intra-rater reliability of the temporal analysis ranged
from r = 0.99 to r = 1.00 for the non-clinical and clinical groups. All the correlations were
significant (p < 0.05). The correlation values for spatial and temporal across the clinical and
non-clinical are presented in Table 6.

92
Table 6. Pearson Product Moment Correlation values depicting the intra-judge measurement
reliability for the spatial and temporal analysis across non-clinical and clinical groups.

Measure Pearson Product Moment Correlation values*

Spatial analysis
Non-clinical group
PPC 0.95
PKC 0.97
Clinical group
PPC 0.95
PKC 0.99

Temporal analysis
Non-clinical group
PPC 0.99
PKC 1.00
Clinical group
PPC 1.00
PKC 1.00

*
All the correlation values were significant (p < 0.05)

93
 Chapter 4. Results
The results are presented in four sections. The first section reports the results of spatial
analysis for the non-clinical and clinical groups. The second section reports the results of the
temporal analysis for the non-clinical and clinical groups. The third section deals with the
effect of age on spatial and temporal learning of speech and keyboard learning tasks in the
non-clinical group. The fourth and final section reports the results of the NASA task load
index.

Spatial learning
Non-clinical group - The results of the spatial analysis for speech and non-speech
(keyboard) learning tasks are shown in Table 7. The speech and keyboard tasks are indicated
in terms of PPC and PKC, respectively. The mean PPC values ranged from 77.5% to 91.6%
across the four practice conditions. The mean PKC values ranged from 59.3% to 96.1%
across the four practice conditions. To evaluate the participant performance on the speech and
keyboard tasks across the four practice conditions, a two-way mixed model analysis of
variance (ANOVA) (2 practice tasks X 4 practice conditions) was performed. The between
group factor was the practice condition (constant, variable, random, & blocked practice
conditions) and the within-group factor was the practice task (speech & keyboard tasks).
Results revealed that there was a significant main effect for practice condition, F (3, 76) =
11.52, p = 0.004, ηp2 = 0.313. Post hoc analysis using Tukey HSD criterion revealed that the
constant practice condition was significantly better than the random (p = 0.007) and blocked
practice conditions (p < 0.001). The variable practice condition was also significantly better
than random (p = 0.034) and blocked practice conditions (p < 0.001). There was a significant
main effect for the practice task with the speech task showing better performance than the
keyboard task, F(1, 76) = 8.632, p = 0.004, ηp2 = 0.102. There was a significant interaction
between the practice conditions and the practice tasks, F(3, 76) = 9.70, p < 0.001,
ηp2 = 0.277. Follow-up post hoc tests for the interaction effect revealed that in the speech
task, constant practice condition was better than the blocked practice condition (p = 0.049),
and random practice condition revealed a marginal significance over blocked practice
conditions (p = 0.06). In the keyboard task, constant practice condition was better than
random (p < 0.001) and blocked practice conditions (p < 0.001), variable practice condition
was also better than random (p < 0.001) and blocked practice conditions (p < 0.001). The
PPC and PKC values for speech and keyboard tasks, respectively across the four practice
conditions are depicted in Figure 13.

94
 Table 7. PPC and PKC values (%) of the participants in the non-clinical group for the speech and non-speech learning tasks. 20 participants were
assigned to each practice condition. Mean (M) and standard deviation values (SD) are shown at the bottom of the table.

Participants Constant Variable Random Blocked

(N = 80) (N = 20) (N = 20) (N = 20) (N = 20)

Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC)

1 93 100 96 100 100 100 100 100

2 79 100 100 100 100 100 82 28.4
3 93 100 100 100 69.48 2.4 93 43.2
4 96 100 91.4 100 82 100 69 38.4
5 83 100 100 100 69 25.8 95.4 100
6 90 100 96.4 100 86 82.6 29 75
7 100 92 69.4 42 85.2 100 61 26.6
8 93 100 100 100 100 29.8 30 46.8
9 100 96.6 93 100 96 73.4 100 54.4
10 100 100 86 100 93 100 100 66.6
11 80.6 100 89 100 75 71.6 100 100
12 96 100 64 100 100 34.2 68 25
13 93 100 96 100 93 14.2 86 42
14 54 100 89 100 100 100 55 34.6
15 100 100 88.8 80 89 100 100 100
16 89.4 100 79 100 100 38.2 96 100
17 100 56.8 42.2 100 89 35.4 33.2 44.8
18 100 100 99.4 83.2 100 58 100 46.4
19 100 100 82 100 99.2 46.8 61 69.2
20 93 78.4 75 100 97 53 93 45.2

M 91.6 96.1 86.8 95.2 91.1 63.2 77.5 59.3

SD 11 10.5 14.9 13.7 10.4 33.4 25.1 27.3

95
 100
90
80
Percentage Correct

70
60
50
PPC
40
30 PKC
20
10
0
Constant Variable Random Blocked
Practice conditions

Figure 13. Percentage of Phoneme Correct (PPC) and Percentage of Keystrokes Correct (PKC) values for
speech and keyboard tasks across four practice conditions in the non-clinical group. Error bars show 95%
confidence interval.

96
 Clinical group - The results of the spatial analysis for speech and non-speech-motor
learning tasks for the clinical group are shown in Table 8. The mean PPC values ranged from
72.5% to 78.1% across the four practice conditions. The mean PKC values ranged from
49.55% to 82.2% across the four practice conditions. The PPC and PKC values were
subjected to a two-way mixed model ANOVA (2 practice tasks X 4 practice conditions) to
determine the participant performance on the speech and keyboard tasks across the four
practice conditions. Results revealed that there was no main effect for the practice condition,
F(3, 12) = 1.325, p = 0.312, ηp2 = 0.249. There was no main effect for the practice task,
F(1, 12) = 1.759, p = 0.209, ηp2 = 0.128. There was also no interaction effect,
F(3, 12) = 1.297, p = 0.320, ηp2 = 0.245.

Non-clinical vs. clinical

Speech task - The mean PPC values for the non-clinical and clinical groups are shown in
Table 9. The mean PPC across the four practice conditions was higher in the non-clinical
group in comparison to the clinical group. In the non-clinical group, the mean PPC was
highest for the constant practice condition (91.6%) and lowest for the blocked practice
condition (77.6%). In the clinical group, the mean PPC was highest for the constant practice
condition (78.1%) and lowest for the variable practice condition (72.5%). To determine the
effect of group (non-clinical vs. clinical) on performance of the speech task, the PPC values
across the four practice conditions were subjected to a two-way ANOVA (2 groups X 4
practice conditions). There was a main effect for group, F(1, 48) = 5.658, p = 0.02, ηp2 = 0.06.
There was no main effect for the practice condition, F(3, 48) = 0.502, p = 0.682, ηp2 = 0.017,
and there was no significant interaction between the group and the practice condition,
F(3, 48) = 0.641, p = 0.682, ηp2 = 0.021.

Keyboard task - The mean PKC for the non-clinical and clinical groups are shown in
Table 9. The mean PKC across all the four practice conditions was higher in the non-clinical
group in comparison to the clinical group. In the non-clinical group, the mean PKC was
highest for the constant practice condition (96.1%) and lowest for the blocked practice
condition (59.3%). The mean PKC for the clinical group was highest for the constant practice
condition (82.2%) and lowest for the blocked practice condition (49.5%).The PKC values
across the four practice conditions were subjected to a two-way ANOVA (2 groups X 4
practice conditions) to determine the effect of group on spatial learning of the keyboard task.
There was a main effect for group, F(1, 48) = 3.925, p = 0.049, ηp2 = 0.043. There was also a

97
 Table 8. PPC and PKC values (%) of the participants in the clinical group for the speech and keyboard tasks across four practice conditions. Four
participants were assigned to each practice condition. Mean (M) and standard deviation (SD) values are shown at the bottom of the table.

Participants Constant Variable Random Blocked

(N = 16) (N = 4) (N = 4) (N = 4) (N = 4)

Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC)

1 40.2 72.2 71 81.6 52.8 46.8 83.6 55

2 92.2 100 75.2 68 60.6 96.6 100 44.8
3 96 56.6 83 100 90 36.8 62.4 25
4 84.2 100 61 61.6 96 36.8 64.2 73.4

M 78.1 82.2 72.5 77.8 74.8 54.2 77.5 49.5

SD 25.7 21.5 9.1 17 21.3 28.6 17.7 20.2

98
Table 9. Mean PPC and PKC values (%) for non-clinical and clinical groups. The standard deviation values are indicated in parentheses.

Groups Constant Variable Random Blocked

Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC) Speech (PPC) Keyboard (PKC)

Non- 91.6 (11) 96.1 (10.5) 86.8 (14.9) 95.2 (13.7) 91.1 (10.4) 63.2 (33.4) 77.5 (25.1) 59.3 (27.3)
Clinical

Clinical 78.1 (25.7) 82.2 (21.5) 72.5 (9.1) 77.8 (17) 74.8 (21.3) 54.2 (28.6) 77.55 (17.8) 49.5 (20.2)

99
main effect for practice condition, F(3, 48) = 8.210, p < 0.001, ηp2 = 0.219. Post hoc analysis
using Tukey HSD criterion revealed that the constant practice condition was better than the
random (p < 0.001) and blocked practice conditions (p < 0.001). The variable practice
condition was also better than random (p < 0.001) and blocked practice conditions
(p < 0.001). There was no significant interaction between the group and the practice
condition, F(3, 48) = 0.343, p = 0.794, ηp2 = 0.021.

Summary of key findings for the spatial learning

Non-clinical group
The retention performance on the speech task was better than on the keyboard task.
The retention performance of the participants in different practice conditions was
influenced by the type of task practised (i.e. there was an interaction between practice
task and practice conditions).
In the speech task, constant practice condition was better than blocked practice
condition. In the keyboard task, constant and variable practice conditions were better
than random and blocked practice conditions.

Clinical group
There was no difference in retention performance between the speech and keyboard
tasks, or between the four practice conditions, and the retention performance of the
participants in different practice conditions was not influenced by the practice task (i.e.
no interaction effect).

Non-clinical vs. clinical group

For the speech task, the retention performance of the non-clinical group was better
than the clinical group, there was no difference in retention performance between the
four practice conditions, and the retention performance of the participants in different
practice conditions was not influenced by the practice group (i.e. clinical or non-
clinical group).
For the keyboard task, the retention performance of the non-clinical group was better
than the clinical group, the performance of participants (in both groups) in the
constant and variable practice conditions was better than random and blocked practice
conditions, and the retention performance of the participants in different practice
conditions was not influenced by the practice group (i.e. clinical or non-clinical).

100
Temporal learning
Non-clinical group - The results of the temporal analysis for speech and keyboard tasks
are shown in Table 10. The results are indicated in terms of phi correlation values. The mean
phi correlation values for speech task ranged from 0.21 to 0.34 across the four practice
conditions. The mean phi correlation values for keyboard task ranged from 0.16 to 0.27
across the four practice conditions. To evaluate the participant performance on the speech and
keyboard tasks, the phi correlation values across the four practice conditions were subjected
to a two-way mixed model ANOVA (2 practice tasks X 4 practice conditions). The between-
group factor was the practice condition (constant, variable, random & blocked practice
conditions) and the within-group factor was the practice task (speech & keyboard tasks).
Results revealed that there was no significant main effect for the practice task,
F(1, 76) = 0.341, p = 0.56, ηp2 = 0.004. There was a significant main effect for the practice
condition, F(3, 76) = 2.901, p = 0.04, ηp2 = 0.103. Post hoc tests using Tukey HSD criterion
revealed that the constant practice condition was significantly better than the random practice
condition (p = 0.03). There was no significant interaction between the practice conditions and
the practice tasks, F(3, 76) = .986, p = 0.40, ηp2 = 0.37. The phi correlation values for speech
and keyboard tasks across four practice conditions are shown in Figure 14.
Clinical group - The results of the temporal analysis for the clinical group indicated in phi
correlation values are shown in Table 11. The mean phi correlation values for speech task
ranged from 0.05 to 0.13 across the four practice conditions. The mean phi correlation values
for the keyboard task ranged from 0.10 to 0.24 across the four practice conditions. To
evaluate the participant performance on the speech and keyboard tasks across the four
practice conditions, a two-way mixed model ANOVA (2 practice tasks X 4 practice
conditions) was performed. There was no main effect for the practice task, F(1, 12) = 3.305,
p = 0.094, ηp2 = 0.189, and no significant main effect for the practice condition,
F(3, 12) = 0.612, p = 0.620, ηp2 = 0.116. There was also no significant interaction between
the practice conditions and practice tasks, F(3, 12) = 0.363, p = 0.781, ηp2 = 0.055.

101
 Table 10. Phi correlation values of the participants in the non-clinical group for the speech and keyboard tasks across four practice conditions. 20
practice conditions were assigned to each practice condition. Mean (M) and standard deviation (SD) values are shown at the bottom of the table

Participants Constant Variable Random Blocked

(N = 80) (N = 20) (N = 20) (N = 20) (N = 20)

Speech (Phi) Keyboard (Phi) Speech (Phi) Keyboard (Phi) Speech (Phi) Keyboard (Phi) Speech (Phi) Keyboard (Phi)

1 0.18 0.49 0.42 0.29 0.12 0.35 0.41 -0.06

2 0.43 0.58 0.49 0.41 0.19 0.57 0.20 0.33
3 0.22 0.00 0.41 0.47 0.35 0.57 0.45 0.51
4 0.49 0.39 0.36 0.16 0.06 0.08 0.14 0.18
5 0.23 0.31 0.36 0.47 0.04 0.08 0.53 -0.08
6 0.00 0.37 0.30 0.03 0.48 0.06 0.30 0.00
7 0.22 0.10 0.55 0.42 0.34 -0.05 0.49 -0.03
8 0.08 0.43 0.22 0.11 0.13 0.28 0.12 0.01
9 0.46 0.17 0.17 0.27 0.33 0.13 0.26 0.47
10 0.40 0.52 0.30 0.54 0.39 -0.10 0.23 0.34
11 0.45 0.16 -0.12 0.04 0.02 0.10 -0.06 0.09
12 0.38 0.00 0.08 0.11 0.17 -0.04 -0.02 0.51
13 0.49 0.24 0.03 0.18 0.27 0.10 0.14 0.21
14 0.49 0.24 -0.04 0.06 0.42 -0.02 0.46 0.29
15 0.00 0.05 0.21 0.32 0.15 0.32 -0.03 0.19
16 0.51 0.28 0.35 0.46 0.28 0.15 -0.03 0.19
17 0.61 0.16 0.05 0.51 0.28 0.27 0.17 0.07
18 0.31 0.31 0.15 0.08 -0.11 0.08 0.00 0.15
19 0.30 0.44 0.28 0.26 0.36 0.19 0.40 0.38
20 0.52 0.23 -0.27 0.30 -0.04 0.11 0.25 0.46

M 0.34 0.27 0.21 0.27 0.21 0.16 0.22 0.21

SD 0.18 0.17 0.21 0.17 0.16 0.19 0.19 0.19
102
 0.4
0.35
0.3
Phi Correlation

0.25
0.2
Speech
0.15
Keyboard
0.1
0.05
0
Constant Variable Random Blocked
Practice Conditions

Figure 14. Phi correlation values of the non-clinical group for speech and keyboard tasks across the four practice
conditions. Error bars show 95% confidence interval.

103
Table 11. Phi correlation values of the participants in the clinical group for the speech and keyboard tasks across four practice conditions. Four
participants assigned to each practice condition. Mean (M) and standard deviation (SD) values are shown at the bottom of the table.

Participants Constant Variable Random Blocked

(N = 16) (N = 4) (N = 4) (N = 4) (N = 4)

Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi)

1 0.14 0.06 0.33 0.28 0.04 0.12 0.27 0.25

2 0.04 0.28 0.14 0.21 0.05 0.01 0.24 0.31
3 0.15 0.27 -0.05 0.31 -0.01 0.18 -0.17 0.05
4 0.17 0.02 0.01 0.16 0.14 0.09 0.03 0.00

M 0.13 0.16 0.10 0.24 0.05 0.10 0.09 0.15

SD 0.06 0.14 0.17 0.07 0.06 0.07 0.20 0.15

104
Non-clinical vs. clinical
Speech task - The mean phi correlation values of speech and keyboard tasks for the non-
clinical and clinical groups are shown in Table 12. As seen from Table 12, the mean
correlation values for the speech task were higher across all four practice conditions in the
non-clinical group in comparison to the clinical group. In the non-clinical group, the mean
correlation value was highest for the constant practice condition (0.34) and lowest for the
variable practice (0.21) and random practice conditions (0.21). The mean correlation value
for the clinical group was highest for the constant practice condition (0.13) and lowest for the
random practice condition (0.05). To determine the effect of group on temporal learning of
the speech task, the phi correlation values across the four practice conditions were subjected
to a two-way ANOVA (2 groups X 4 practice condition). Results revealed a main effect for
group with the non-clinical group performance being better than the clinical group, F(1, 48) =
4.534, p = 0.038, ηp2 = 0.086. There was no significant main effect for practice condition,
F(3, 48) = 2.479, p = 0.072, ηp2 = 0.134. There was no significant interaction between group
performance and practice conditions, F(3, 48) = 1.680, p = 0.184, ηp2 = 0.095.

Keyboard task - From Table 12, it is suggestive that the phi correlation values in the non-
clinical group were highest for the constant and variable practice conditions (0.27) and lowest
for the random practice condition (0.16). In the clinical group, the variable practice condition
revealed the highest correlation value (0.24) and the random practice had the lowest
correlation value (0.10). The phi correlation values across the four practice conditions were
subjected to a two-way ANOVA (group X practice condition). There was no significant main
effect for group, F(1, 48) = 0.704, p = 0.405, ηp2 = 0.014. There was no significant main
effect for practice condition, F(3, 48) = 0.804, p = 0.498, ηp2 = 0.048. There was also no
significant interaction between the groups and practice conditions, F(3, 48) = 0.697,
p = 0.559, ηp2 = 0.042.

105
Table 12. Mean phi correlation values for the non-clinical and clinical groups. The standard deviation values are shown in parentheses

Group Constant Variable Random Blocked

Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi) Speech (phi) Keyboard (phi)

Non-clinical 0.34 (0.18) 0.27 (0.17) 0.21 (0.21) 0.27 (0.17) 0.21 (0.16) 0.16 (0.19) 0.22 (0.19) 0.21 (0.19)

Clinical 0.13 (0.06) 0.16 (0.14) 0.10 (0.17) 0.24 (0.17) 0.05 (0.06) 0.10 (0.07) 0.09 (0.20) 0.15 (0.15)

106
Summary of the key findings for the temporal learning
Non-clinical group findings
There was no difference in retention performance between the speech and keyboard
task.
The retention performance in constant practice condition was better than the random
practice condition. There was no difference between the variable, random, and blocked
practice conditions.
The retention performance of the participants in different practice conditions was not
influenced by the practice task (i.e., no interaction effect).
Clinical group findings
There was no difference in retention performance between the speech and keyboard
task, no difference between the practice conditions, and the retention performance of
the participants in different practice conditions was not influenced by the practice task.
Non-clinical vs. clinical group findings
For the speech task, the retention performance of the non-clinical group was better
than the clinical group, the retention performance of the participants in the constant
practice condition was better than the random practice condition, and the retention
performance of the participants in different practice conditions was not influenced by
the practice group (i.e., no interaction effect).
For the keyboard task, there was no difference in the retention performance between
the non-clinical and clinical groups, no difference between the four practice
conditions, and the retention performance of the participants in different practice
conditions was not influenced by the practice group (i.e., no interaction effect).

107
Age effect
The effect of age on spatial and temporal learning was analysed in the non-clinical group.
A similar analysis was not performed in the clinical group due to inadequate sample size. The
age of the participants in the non-clinical group ranged from 42 to 78 years. To examine the
effect of age on motor learning performance, a median split was performed to separate the
participants into two groups (younger, older). The median split was at 59 years of age, thus
placing 40 participants in each age group. The age distribution across the four practice
conditions included nine younger participants in the constant practice condition, 11 younger
participants in the variable practice conditions, and 10 younger participants each in random
and blocked practice conditions. The distribution of the participants across the four practice
conditions for the spatial learning and temporal learning tasks is depicted in tables 13 through
16. The age of the participants in the younger group ranged from 42 to 59 years (M = 52.2),
and in the older group, the age of the participants ranged from 60 to 79 years (M = 67.3).

Spatial learning
Speech task - The mean PPC and PKC values for the younger and older age groups are
shown in Table 17. The mean PPC values across four practice conditions are higher in the
younger age group except for the constant practice condition. In the younger group, the mean
PPC was highest for the random practice condition (93.5%) and lowest for the constant and
variable practice conditions (90.1%). In the older group, the mean PPC was highest for the
constant practice condition (92.9%) and lowest for the blocked practice condition (63.8%).
To determine the age effect on speech task, the PPC values across the four practice conditions
were subjected to a two-way ANOVA (2 age groups X 4 practice conditions). There was a
significant main effect for age with the younger group performing better than the older group,
F(1, 72) = 7.390, p = 0.008, ηp2 = 0.093. A significant main effect or the practice condition
was also found, F(3, 72) = 3.674, p = 0.016, ηp2 = 0.133. Post hoc analysis using Tukey HSD
criterion revealed that the constant practice condition was better than the blocked practice
condition (p = 0.022). There was an interaction between the age of the participants and the
practice conditions, F(3, 72) = 3.657, p = 0.016, ηp2 = 0.132. Follow-up post hoc tests for the
interaction effect revealed that in the older group, the retention performance participants in
the constant practice condition was significantly better than the participants in the blocked
practice condition (p = 0.01). In the younger group, there was no difference in the retention
performance between the participants in the four practice conditions. The PPC values of the
younger and older age groups across four practice conditions are depicted in Figure 15.

108
 Table 13. Distribution of Percentage of Phoneme Correct values (%) in the younger and older
age groups for the speech task across four practice conditions. 2 Mean (M) and Standard
Deviation (SD) values are shown at the bottom of the table.

Participants Constant Variable Random Blocked

Young Old Young Old Young Old Young Old

1 93 93 86 100 100 69.48 100 69

2 100 96 93 91.4 85.2 86 86 100
3 89.4 100 96.4 89 93 69 100 68
4 80.6 54 64 100 100 93 100 29
5 93 90 96 69.4 100 99.2 61 55
6 79 100 89 96 100 100 93 61
7 83 100 99.4 75 100 96 95.4 93
8 93 100 100 82 82 97 82 30
9 100 100 79 42.2 100 89 96 33.2
10 96 88.8 75 89 100 100
11 93 100

M 90.1 92.9 90.1 82.7 93.5 88.7 91.3 63.8

SD 7.7 13.3 10.9 18.6 9.4 11.2 12.3 27.6

2
Unequal sample sizes across the four groups is a result of median split performed on the overall data

109
Table 14. Distribution of percentage of keystrokes correct values (%) in the younger and
older age groups for the keyboard task across four practice conditions. Mean (M) and
Standard Deviation (SD) values are shown at the bottom of the table.

Participants Constant Variable Random Blocked

Young Old Young Old Young Old Young Old

1 100 100 100 100 100 2.4 100 38.4

2 100 100 100 100 100 82.6 42 100
3 100 92 100 100 14.2 25.8 100 25
4 100 100 100 100 29.8 100 54.4 26.6
5 100 100 100 42 58 46.8 43.2 34.6
6 100 96.6 100 100 100 38.2 69.2 75
7 100 56.8 83.2 100 100 53 100 45.2
8 100 100 100 100 34.2 73.4 28.4 44.8
9 100 100 100 100 100 35.4 100 46.8
10 100 80 71.6 100 66.6 46.4
11 78.4 100

M 100 93 96.6 93.5 70.7 55.7 70.3 48.2

SD 0 13.7 7.4 19.3 34.4 32.4 28.0 22.8

110
Table 15. Distribution of phi correlation values in the younger and older age groups for the
speech task across four practice conditions. Mean (M) and Standard Deviation (SD) values
are shown at the bottom of the table.

Participants Constant Variable Random Blocked

Young Old Young Old Young Old Young Old

1 0.18 0.40 0.42 0.08 0.12 0.02 0.41 -0.06

2 0.43 0.45 0.49 0.03 0.19 0.17 0.20 -0.02
3 0.22 0.38 0.41 -0.04 0.35 0.27 0.45 0.14
4 0.49 0.49 0.36 0.21 0.06 0.42 0.14 0.46
5 0.23 0.49 0.36 0.35 0.04 0.15 0.53 -0.03
6 0.00 0.00 0.30 0.05 0.48 0.28 0.30 -0.03
7 0.22 0.51 0.55 0.15 0.34 0.28 0.49 0.17
8 0.08 0.61 0.22 0.28 0.13 -0.11 0.12 0.00
9 0.46 0.31 0.17 -0.27 0.33 0.36 0.26 0.40
10 0.30 0.30 0.39 -0.04 0.23 0.25
11 0.52 -0.12

M 0.26 0.41 0.31 0.09 0.24 0.18 0.31 0.13

SD 0.17 0.16 0.18 0.19 0.15 0.18 0.15 0.19

111
Table 16. Distribution of phi correlation values in the younger and older age groups for the
keyboard task across four practice conditions. Mean (M) and Standard Deviation (SD) values
are shown at the bottom of the table.

Participants Constant Variable Random Blocked

Young Old Young Old Young Old Young Old

1 0.49 0.52 0.29 0.11 0.35 0.10 -0.06 0.09

2 0.58 0.16 0.41 0.18 0.57 -0.04 0.33 0.51
3 0.00 0.00 0.47 0.06 0.57 0.10 0.51 0.21
4 0.39 0.24 0.16 0.32 0.08 -0.02 0.18 0.29
5 0.31 0.24 0.47 0.46 0.08 0.32 -0.08 0.19
6 0.37 0.05 0.03 0.51 0.06 0.15 0.00 0.19
7 0.10 0.28 0.42 0.08 -0.05 0.27 -0.03 0.07
8 0.43 0.16 0.11 0.26 0.28 0.08 0.01 0.15
9 0.17 0.31 0.27 0.30 0.13 0.19 0.47 0.38
10 0.44 0.54 -0.10 0.11 0.34 0.46
11 0.23 0.04

M 0.32 0.24 0.29 0.25 0.20 0.12 0.17 0.25

SD 0.19 0.15 0.18 0.16 0.24 0.11 0.23 0.15

112
 Table 17. PPC and PKC values (%) of the younger and older age groups across four practice
conditions for speech and keyboard learning tasks. The standard deviation values are shown
in parentheses.

Age groups Constant Variable Random Blocked

Speech Keyboard Speech Keyboard Speech Keyboard Speech Keyboard

(PPC) (PKC) (PPC) (PKC) (PPC) (PKC) (PPC) (PKC)

Young 90.1 100 90.1 96.6 93.5 70.7 91.3 70.3

(7.79) (0) (10.91) (7.48) (9.41) (34.40) (12.37) (28.06)

Old 92.9 93.1 82.7 93.5 88.7 55.7 63.8 48.2

(13.39) (13.71) (18.60) (19.33) (11.27) (32.49) (27.68) (22.88)

113
 120

100

80
PPC

60
Younger
40 Older

20

0
Constant Variable Random Blocked
Practice Conditions

Figure 15. Percentage of Phoneme Correct (PPC) values of the younger and older age groups across four
practice conditions. Error bars show 95% confidence interval.

114
 Keyboard task - The mean PKC across all of the practice conditions was higher in the
younger group in comparison to the older group. In the younger group, the mean PKC was
highest for the constant practice condition (100%) and lowest for the blocked practice
condition (70.3%). In the older group, the mean PKC was highest for the variable practice
condition (93.5%) and lowest for the blocked practice condition (48.2%).The PKC values
across the four practice conditions were subjected to a two-way ANOVA (2 age groups X 4
practice conditions) to determine the effect of age on keyboard task performance. There was
a significant main effect for age with the younger group performance being better than the
older group, F(1, 72) = 5.24, p = 0.024, ηp2 = 0.069. A significant main effect for the practice
condition was also found, F(3, 72) = 15.337, p = 0.001, ηp2 = 0.390. Post hoc analysis using
Tukey HSD criterion revealed that the constant practice condition was better than random
(p < 0.001) and blocked practice conditions (p < 0.001), and performance on the variable
practice condition was better than random (p < 0.005) and blocked practice conditions
(p <0.001). There was no significant interaction between the age of the participants and the
practice conditions, F(3, 72) = 0.690, p = 0.561, ηp2 = 0.028. The PKC values of the younger
and older age groups across four practice conditions are depicted in Figure 16.

Temporal learning
Speech task - The mean phi correlation values for the younger and older age groups are
shown in Table 18. The mean correlation values were higher in the younger age group in
comparison to the older group across practice conditions except for the constant practice
condition. In the younger group, the mean correlation value was highest for the variable
practice condition (0.31) and lowest for the random practice condition (0.24). In the older
group, the mean correlation value was highest for the constant practice condition (0.41) and
lowest for the variable practice condition (0.09). To determine the effect of age on speech
task performance, the phi correlation values across the four practice conditions were
subjected to a two-way ANOVA (2 age groups X 4 practice conditions). Results revealed a
significant main effect for age with the performance of the younger group better than the
older group, F(1, 72) = 4.352, p = 0.04, ηp2 = 0.057. There was no main effect for the practice
condition, F(3, 72) = 2.426, p = 0.072, ηp2 = 0.092. There was a significant interaction
between the age of the participants and the practice conditions, F(3, 72) = 4.708 , p = 0.005,
ηp2 = 0.164. Follow up post hoc tests revealed that the retention performance of the
participants in the constant practice condition was better than participants in the variable,
random, and blocked practice conditions (p < 0.05) in the older age group but there were no

115
 120

100

80
PKC

60
Younger
40 Older

20

0
Constant Variable Random Blocked
Practice Conditions

Figure 16. Percentage of keystrokes correct (PKC) values of the younger and older age groups across four
practice conditions. Error bars show 95% confidence interval.

116
Table 18. Phi correlation values of the younger and older age groups across four practice conditions
for speech and keyboard learning tasks. The standard deviation values are shown in parentheses.

Age groups Constant Variable Random Blocked

Speech Keyboard Speech Keyboard Speech Keyboard Speech Keyboard

(phi) (phi) (phi) (phi) (phi) (phi) (phi) (phi)

Young 0.26 0.32 0.31 0.29 0.24 0.20 0.31 0.17

(0.17) (0.19) (0.18) (0.18) (0.15) (0.24) (0.15) (0.23)

Old 0.41 0.24 0.09 0.25 0.18 0.12 0.13 0.25

(0.16) (0.15) (0.19) (0.16) (0.18) (0.11) (0.19) (0.15)

117
differences between the practice conditions in the younger group (p = 0.12). The phi
correlation values of the younger and older age groups for the speech task across four
practice conditions is depicted in Figure 17.

0.6

0.5
Phi Correlation

0.4

0.3
Younger
0.2 Older

0.1

0
Constant Variable Random Blocked
Practice Conditions

Figure 17. Phi correlation values of the younger and older age groups for the speech task across four practice
conditions. The error bars show 95% confidence interval.

118
 Keyboard task - The mean phi correlation values of the younger and older groups are
shown in Table 18. The correlation values in the younger group were highest for the constant
practice condition (0.32) and lowest for the blocked practice condition (0.17). In the older
group, variable and blocked practice conditions revealed the highest correlation value (0.25)
and the random practice had the lowest correlation value (0.12).The phi correlation values
across the four practice conditions were subjected to a two-way ANOVA to determine the
effect of age on temporal learning of the keyboard task. There was no significant main effect
for age, F(1, 72) = 0.370, p = 0.545 ηp2 = 0.005. There was also no significant main effect for
practice condition F(3, 72) = 1.841, p = 0.147, ηp2 = 0.071. There was no significant
interaction between the age of the participants and the practice conditions, F(3, 72) = 0.885, p
= 0.453, ηp2 = 0. 036, indicating that the age of the participants did not affect the temporal
learning of the keyboard task across practice conditions. The phi correlation values of the
younger and older age groups for the keyboard task across four practice conditions is
depicted in Figure 18.

119
 0.5
0.45
0.4
0.35
Phi Correlation
0.3
0.25
Younger
0.2
0.15 Older

0.1
0.05
0
Constant Variable Random Blocked
Practice Conditions

Figure 18. Phi correlation values of the younger and older age groups for the keyboard task across four practice
conditions. The error bars show 95% confidence interval.

120
Summary of key findings of age effect on spatial and temporal learning
Speech task
The retention performance of the younger group was better than the older group in
spatial as well as temporal learning.
The retention performance of the participants in different practice conditions was
influenced by the age group for spatial and temporal learning (i.e. there was an
interaction effect).
In case of spatial learning, follow up post hoc tests revealed that the retention
performance of the participants in the constant practice condition was better than the
participants in the blocked practice condition in the older age group but there were no
differences between the practice conditions in the younger group.
In case of temporal learning, follow up post hoc tests revealed that the retention
performance of the participants in the constant practice condition was better than
participants in the variable, random, and blocked practice conditions in the older age
group, but there were no differences between the practice conditions in the younger
group.

Keyboard task
The retention performance of the younger group was better than the older group in
spatial but not in temporal learning.
The retention performance of the participants in different practice conditions was not
influenced by the age group for spatial as well as temporal learning (i.e. there was no
interaction effect).
In case of spatial learning, the retention performance (of the younger and older age
group participants) was better in constant and variable practice conditions than
random and blocked practice conditions. In case of temporal learning, there was no
difference between the practice conditions.

121
NASA Task Load Index
Each participant completed the NASA task load index to evaluate the perceived difficulty
related with performance on speech and keyboard tasks. The effect of age on perceived task
difficulty was analysed in the non-clinical group alone. A similar analysis was not performed in
the clinical group due to inadequate sample size. The splitting of participants into younger and
older age groups was performed in a similar manner used to investigate the effect of age on
speech and keyboard learning tasks. The age distribution of the participants across four practice
conditions included nine younger participants in the constant practice conditions, 11 younger
participants in variable practice conditions, and 10 younger participants each in random and
blocked practice conditions.

Speech task
The NASA task load index scores of the younger and older age groups across the four
practice conditions is shown in Figure 19. The mean index scores across four practice conditions
were higher in the older group except in case of variable practice condition. In the younger age
group, the mean index score was highest for the random practice condition (8.73) and lowest for
the constant practice condition (5.14). In the older age group, the mean index score was highest
for the random practice condition (10.53) and lowest for the constant practice condition (7.58).
To determine the effect of age on the perceived difficulty of the speech task, the mean index for
the four practice conditions were subjected to a two-way ANOVA (2 age groups X 4 practice
conditions). There was a main effect for age, F(1, 72) = 4.737, p = 0.033, ηp2 = 0.062. There was
also a main effect for practice condition, F(3, 72) = 4.784, p = 0.004, ηp2 = 0.166. Post hoc
analysis using Tukey HSD criterion revealed that constant practice condition received lower
score than random (p = 0.006) and blocked practice conditions (p = 0.026). There was no
interaction between the age of the participants and the practice condition, F(3, 72) = 1.002,
p = 0.397, ηp2 = 0.040.

122
 14

12
NASA index scores

10

6 Young
Old
4

0
Constant Variable Random Blocked
Practice conditions

Figure 19. NASA task load index scores of the speech task across four practice conditions for younger and older
groups. The error bars show 95% confidence interval.

123
Keyboard task
The NASA task load index scores of the younger and older age groups across the four
practice conditions is shown in Figure 20. The mean index scores across four practice conditions
are higher in the older group except the blocked practice condition. In the younger age group, the
mean index score was highest for the blocked practice condition (12.1) and lowest for the
constant practice condition (6.27). In the older age group, the mean index score was highest for
the random practice condition (12.55) and lowest for the variable practice condition (7.5). To
determine the effect of age on the perceived difficulty of the keyboard task, the mean index for
the four practice conditions were subjected to a two-way ANOVA (2 age groups X 4 practice
conditions). There was no main effect for age, F(1, 72) = 0.25, p = 0.618, ηp2 = 0.003. However,
there was a main effect for practice condition, F(3, 72) = 11.848, p < 0.001, ηp2 = 0.331. Post hoc
analysis using Tukey HSD criterion revealed that constant practice condition received lower
score than random (p < 0.001) and blocked practice conditions (p < 0.001). There was no
interaction between the age of the participants and the practice condition, F(3, 72) = 1.866,
p = 0.143, ηp2 = 0.072.

Speech vs. keyboard tasks

To compare the perceived difficulty of speech vs. keyboard learning task, the NASA load
index scores across all the four practice conditions were collapsed in both the tasks and
compared using a paired sample ‘t’ test. On average, the index scores of the keyboard task (M =
9.85, SD = 3.25) was higher than the speech task (M = 8.44, SD = 3.52), t(79) = -4.103, p <
0.001 (two-tailed), r = 0.59. The mean index scores for the speech and keyboard tasks are shown
in Figure 21.

124
 16
14
12
NASA index scores

10
8
Young
6
Old
4
2
0
Constant Variable Random Blocked
Practice conditions

Figure 20. NASA task load index scores of the younger and older groups across four practice conditions for the
keyboard task. The error bars show 95% confidence interval.

125
 10.00
NASA index scores

8.00

6.00
Speech
4.00
Keyboard

2.00

0.00
Speech Keyboard
Practice task

Figure 21. Mean NASA task load index scores for the speech and keyboard tasks. Error bars show 95% confidence
interval.

126
Summary of key findings of NASA task load index
Speech task
The older group perceived the speech task to be more difficult than the younger group.
Across both the groups, the constant practice condition was perceived easier in
comparison to random and blocked practice conditions.
The age of the participants did not influence the participants’ performance across the four
practice condition (i.e. no interaction effect).

Keyboard task
There was no difference in the mean index scores between the younger and older groups.
Across both the groups, the constant practice condition was perceived easier than random
and blocked practice conditions.
The age of the participants did not influence the participants’ performance across the four
practice condition (i.e. no interaction effect).

Speech vs. Keyboard tasks

The keyboard task was perceived to be more difficult than the speech task.

127
 Chapter 5. Discussion
The major findings of the current study can be summarised as follows: First, in the case of
spatial learning, performance on the speech task was better than the keyboard task. The constant
practice was found to be better than blocked practice for the speech task. For the keyboard task,
constant and variable practice conditions were better than random and blocked practice
conditions. Second, in the case of temporal learning, there was no significant difference between
speech and keyboard tasks. The main effect of practice condition across speech and keyboard
tasks revealed that the constant practice condition was better than the random practice condition.
Third, an age effect was observed, with the performance of the younger age group being better
than the older age group across all of the speech and keyboard tasks. Fourth, performance of the
non-clinical group was better than the clinical group on both the speech and keyboard tasks. A
discussion of the results in regard to each of the major findings is provided below.

Spatial Learning
Non-clinical group
Speech task - The mean PPC values of the constant, variable, random, and blocked practice
conditions were 91.6%, 86.8%, 91.1%, and 77.5%, respectively, with the constant practice
condition being better than the blocked practice condition. There have been limited studies
investigating the effects of PMLs on speech-motor learning, and in particular there has been no
research investigating the application of PMLs in spatial learning among healthy individuals.
Rosenbek, Lemme, Ahern, Harris, and Wertz (1973) explored the efficacy of the ‘eight-step
continuum approach’ in treating speech deficits in three adults with AOS. The eight-step
continuum approach is based on a hierarchical cueing procedure that begins with a high level of
support providing simultaneous production of slowly spoken simple utterances with visual and
tactile cues. During the course of therapy, these cues are either gradually faded or increased until
the cues are completely faded and the patient begins producing delayed repetitions of
increasingly complex stimulus items. The results revealed that the treatment outcome varied
among the three individuals. Even though the researchers did not mention directly about the
application of PMLs during the treatment protocol, they speculated that constant practice might
facilitate acquisition of new utterances, whereas variable practice might help in retention. The
variable nature of treatment outcome and low sample size in the Rosenbek et al. study makes it

128
difficult to judge the effectiveness of variable practice. The elaboration hypothesis is attributed to
the beneficial nature of variable practice (Shea & Morgan, 1979). An individual engaged in
variable practice elaborates the memory representations of the skill variations of the practising
task, and this helps the individuals to compare and contrast the skills variation and thus
eventually facilitates learning the task. It is likely for this reason that variable practice might
have benefitted the participants in Rosenbek et al.’s study. On the contrary, the present study
found no significant difference between constant and variable practice conditions during the
retention phase. The reason for this discrepancy can be attributed to Challenge Point Framework
(CPF) (Guadagnoli & Lee, 2004). CPF explains the effects of various practice conditions on
motor learning tasks. According to CPF, successful learning of a task depends on the skill level
of the learner and the difficulty of the to-be learned task. With regards to the skill level of the
learner, Rosenbek et al. recruited impaired speakers (AOS) as participants, whereas the present
study examined healthy participants. In terms of the difficulty of the task to-be learned,
Rosenbek at al. used meaningful words and phrases as practice stimuli, whereas the present study
used complex phrase(s) consisting of non-words as practice stimulus. It is plausible that as the
complexity of the practising task increases, the practice conditions which allow the participants
to engage in repeated practice might facilitate the spatial learning of the task.3 The repeated
exposure to the visual and auditory representations of the target phrase offered by the constant
and variable practice conditions could have equally benefited the participants in spatial learning
of the target phrase.
There have been two studies of impaired speakers using random vs. blocked practice
conditions and the results of these studies are equivocal (Knock et al., 2000; Mass & Farinella,
2012). Knock et al. compared random vs. blocked practice in two adults diagnosed to have
aphasia as well as severe AOS, and found both the practice conditions to be equally beneficial
during the acquisition phase of learning. However, during the retention phase both individuals
showed poorer maintenance of blocked practice targets than random practice targets. Mass and
Farinella compared random vs. blocked practice in four children diagnosed with AOS. The
findings during the retention phase were mixed, with blocked practice benefitting two children,
random practice benefitting one

3
The variable practice condition involved practicing the target phrase along with an alternate phrase. Both the
phrases were similar in terms of spatial representation with the only difference being the temporal organization of

129
the phrase. With regards to spatial learning, the participants were practising the same phrase throughout the
acquisition phase, which could have facilitated the formation of an efficient cognitive representation .
child, no clear improvement in either condition was seen in another child.
The difference between the findings of Knock et al. and Mass and Farinella can be attributed
to two reasons. First, is the inherent performance variability seen between individuals with AOS
[American Speech, Language, and Hearing Association (ASHA), 2007)]. The diverse nature of
AOS represents a situation where two individuals with the same severity might exhibit varied
speech deficits and respond differently to the same therapy technique. The second reason might
be due to the age of the participants. Knock et al. found random practice to be beneficial in
adults, whereas Mass and Farinella found blocked practice to be better in children. It is possible
that beneficial effects of practice conditions are different for children than for adults. In the
present study, the random practice condition had a marginal significant advantage (p = 0.06)
over blocked practice condition in terms of retention performance, similar to the findings of
Knock et al. (2000). It can be speculated that as both the studies recruited adult participants,
random practice might benefit adults rather than children in complex speech learning tasks.
Overall, across the four practice conditions, the results reveal that the constant practice was
condition was better than the blocked practice condition. It is likely that the repeated exposure
(as in the case of constant practice) to the orthographic stimulus could have been a cue to the
phonological motor plan that helped in the retrieval of the speech phrase during the retention
phrase. Laguna (2008) proposed that repeated practice and exposure to the visual image of the
practicing task can aid in the development of memory representation of the task, called
‘cognitive representation’. Past research has also proven that auditory feedback can be vital in
learning new speech sounds, as speakers use auditory perceptual features as a reference for
articulation of the novel speech sounds (Perkell et al., 1997). In the current study, the continuous
visual and auditory representations of the target phrase offered by the constant practice could
have facilitated the participants to form an accurate cognitive representation of the phrase.
However, the nature of the blocked practice did not provide continuous visual and auditory
representations of the target phrase as the target and alternate phrases were separated by a block
of 25 trials. This sequential manner of practice could have prevented the participants from
storing the target speech phrase in their working memory leading to memory trace decay. This
could be attributed to the decreased spatial performance of the participants in the blocked
practice condition.

130
 Keyboard task - The mean PKC values of the constant, variable, random and blocked practice
conditions were 96.1%, 95.2%, 63.2%, and 59.3%, respectively. The results indicated that the
constant and variable practice conditions were significantly better than the random and blocked
practice conditions. This finding differs from past research which has found variable practice to
be more beneficial than constant practice for a number of non-speech tasks like basketball
shooting, and target tracking tasks (Shoenfelt at al., 2002; Wulf & Schmidt, 1997). For example,
Shoenfelt et al. compared the beneficial effects of constant vs. variable practice in shooting
basketball free throws. The researchers found that the constant as well as variable practice
groups improved during the acquisition phase. On a retention test after two weeks, the variable
practice group demonstrated much better performance than during the acquisition phase. On the
other hand, the constant practice group, returned to their pre-test level on the retention test.
Similar to past research examining speech-motor learning, non-speech-motor learning is
facilitated by variable practice. Yet, the present study found constant and variable practice
conditions to be equally beneficial for non-speech-motor learning. As in the case of speech task,
it is likely that the repeated exposure to the visual and auditory representations of the target tune
offered by the constant and variable practice conditions could have equally benefited the
participants in spatial learning of the target tune.
With regards to random vs. blocked practice conditions, numerous studies have shown the
benefits of random over blocked practice conditions across a wide range of tasks like throwing
balls at a target, maze tracing, and computer-based tracking (Goode & Magill, 1986; Shea &
Wright, 1991). For example, Goode and Magill found that in throwing a ball to a target, random
practice led to better retention in comparison to blocked practice. However, in the present study,
participants in random, as well as blocked practice conditions demonstrated similar performance.
It is possible that as the participants did not have prior experience in playing the keyboard, they
may have required repeated exposure to the visual and auditory representations of the target tune
without any interruption of the alternate tune. However, the interference of the alternate tune in
the random and blocked practice conditions could have precluded the participants from learning
the target tune successfully. This can possibly account for the similar performance of the
participants in the random and blocked practice conditions.

131
 Speech vs. Keyboard tasks - Examination of the PPC and PKC scores revealed that the
participants performed better on the speech task compared to the keyboard task. This can be
attributed to two possibilities. First, the speech task offered an orthographic representation of the
target speech phrase during the practice regime. Previous research has shown that it is easier to
create a mental image of lexical items than non-lexical items (Prado & Ullman, 2009). Second,
as speech is a highly practised task, it is possible that the auditory feedback could have helped
the participants to develop a better internal model of the target phrase in comparison to the
keyboard task. The internal model is a representation of the articulatory configurations
associated with various sounds produced in the vocal tract (Perkell et al., 2000). In contrast, the
participants were not accustomed to the keyboard task, and this lack of familiarity could have
precluded the participants from developing an internal model of the keyboard task even in the
presence of auditory feedback.
The initial proposed hypothesis was “The PMLs that best facilitate spatial learning of a novel
musical keyboard entry task (non-speech task) will also best facilitate spatial learning of a novel
speech utterance (speech task) in healthy individuals”. The findings of the spatial learning
partially support this hypothesis. Across the speech and keyboard tasks, constant practice
condition provided maximum retention whereas blocked practice conditions offered the least
retention. On the whole, the findings of the spatial learning in the current study suggest that
constant practice condition might prove to be beneficial in learning complex speech and non-
speech-motor tasks.

Temporal Learning
Non-clinical group
Speech task - The phi correlation values for constant, variable, random, and blocked practice
conditions were 0.34, 0.21, 0.21, and 0.22, respectively. Statistical analyses found no significant
difference between the four practice conditions, although the highest correlation was found
among the participants in the constant practice condition.
Adams and Page (2000) trained participants to practise the utterance “Buy Bobby a poppy”
with a specific overall utterance duration, using either constant or variable practice conditions.
Results of the training revealed that the group undergoing constant practice were less successful

132
in their training compared to the variable practice group, despite the fact that the constant group
had received twice as many practice trials as the variable practice group. These results indicated
that the use of variable practice was more beneficial than constant practice for temporal learning
of speech in unimpaired speakers.
The notion of ‘speed-accuracy trade-off’ (SAT) is commonly implicated in motor learning
tasks which demands spatial and temporal accuracy (Wickelgren, 1977; Dickman & Meyer,
1988; Jelsma & Pieters, 1989). In a typical SAT situation, the speed of the motor skill is reduced
when focus is on accuracy and vice-versa (Schmidt & Lee, 2005). According to Fitts (1954),
there is a proportional relationship between movement time and accuracy. It means that as speed
increases, accuracy decreases proportionally. In the speech-motor learning literature, the SAT
situation has been reported in various studies (Gooze et al., 2005; Parnell & Amerman, 1996;
Amerman & Parnell, 1990). In a recent study, Latash and Mikaelian (2011) explored the
relationship between task difficulty and speech time in picture description tasks. They found that
speech time scaled linearly with the increase in difficulty of the naming tasks. This was termed
as the speed-difficulty trade-off situation rather than a typical speed-accuracy trade-off situation.
In the present study, the speed in performing the motor tasks was not measured, but rather the
temporal performance was measured via phi correlation approach. The low correlation values
and similar temporal performance across the four practice conditions is suggestive of a decreased
temporal learning by the participants, and this can be attributed to the ‘spatial-temporal trade-off’
similar to the notion of SAT. It is plausible that as the participants had to learn a novel speech
task, they would have focused more on spatial accuracy, thus compromising temporal accuracy.

Keyboard task - The phi correlation values for constant, variable, random, and blocked
practice conditions were 0.27, 0.27, 0.16, and 0.21, respectively, with no significant difference in
the phi correlation values across the four practice conditions. Previous studies have investigated
the effect of various practice conditions on learning absolute and relative timing (Shea, Lai,
Wright, Immink, & Black, 2001; Sekiya et al., 1996). The results of the past studies suggest that
absolute timing is enhanced by random practice and relative timing is enhanced by constant and
blocked practice conditions. For example, Shea et al. (2001) compared the effects of constant,
serial (practice condition in which the task variation change from trial to trial in a predictable
manner), random, and blocked practice conditions on learning absolute and relative timing in a
sequential keyboard pressing task. Results revealed that participants who were in the constant

133
and blocked practice conditions demonstrated better relative timing abilities, whereas better
absolute timing abilities were demonstrated by participants in the random and serial practice
conditions. The researchers explained the results based on the predictability of the practice
environment. Predictability of the practice environment refers to the ease with which the
participants can predict the forthcoming task variations during the practice regime (e.g., a
blocked practice conditions offers a highly predictive practice environment). Practice conditions
which facilitated predictability of the forthcoming tasks helped in the learning of relative timing,
whereas practice conditions which enhanced variability and unpredictability of the forthcoming
tasks helped in learning of absolute timing.
There are two methodological differences that make it difficult to directly compare the results
of the current study with the previous studies. First, the present study investigated temporal
learning through the calculation of a phi correlation using absolute duration measures. In
contrast, past studies have directly measured the absolute duration and relative duration values of
individual motor movements (e.g., duration of a speech utterance). Second, the focus of past
studies was restricted to temporal learning. In the present study, the participants were asked to
focus on both spatial as well as temporal learning.

Speech vs. keyboard tasks - There was no main effect of the practice task, indicating that
temporal learning was similar across the speech and keyboard tasks. The low phi correlation
values on both the speech and keyboard tasks indicate that participants had considerable
difficulty in synchronizing their productions with the target phrase/tune during the retention
trials. There are two possible reasons for this.
First, is the possibility of a ‘spatial-temporal trade-off’. The participants were required to learn
the spatial as well as temporal aspects of speech and keyboard tasks. It is likely that when
participants are learning tasks which are complex and novel in nature (as in the present study),
they might tend to focus more on performing those tasks correctly with respect to the spatial
domain rather than focussing on the temporal domain. Even in the present study, during the
retention phase the participants could have been more attentive in saying the words correctly or
playing the correct keys rather than timing their productions to match the temporal duration of
the target phrase/tune.
The second reason could be attributed to the validity of the phi correlation approach to
assessing the temporal learning. The current method of evaluating temporal learning did not

134
parallel the approach used in previous studies, namely measurement of relative duration, syllable
length, and pause duration. Instead, a broader approach was adopted by measuring the overall
match between the participants’ productions and target stimulus. As there have been no prior
studies investigating the use of phi correlation in estimating motor learning, it is likely that this
novel approach may be insensitive to certain subtle features of temporal learning that may have
been present.
The initial proposed hypothesis was “The PMLs that best facilitate temporal learning of a
novel musical keyboard entry task (non-speech task) will also best facilitate temporal learning of
a novel speech utterance (speech task) in healthy individuals”. This hypothesis can either be
supported or rejected based on the interpretation of the current findings. Since there was no
significant difference between the four practice conditions (in terms of retention) in speech as
well as keyboard tasks, one line of interpretation could be that all the PMLs were equally
facilitative in temporal learning of the keyboard entry as well as the speech task, thus providing
support for this hypothesis. However, the current findings can also be interpreted in a different
manner. As there was no significant difference between the four practice conditions across both
the tasks, none of the PMLs best facilitated the temporal learning of the keyboard entry as well
as the speech task, thus offering ground for rejecting this hypothesis.

Age effect
Spatial learning
Speech task - Based on a median split of the data according to age of participants, the younger
age group performed better than the older age group across the four practice conditions. The PPC
values across the four practice conditions indicated a similar performance among the young
participants. In the older age group, the constant practice condition was better than the blocked
practice condition. Considered within the context of the overall results, it appears that the
difference observed in spatial learning across the practice conditions were primarily found
among the older participants.
The poorer performance found among the older participants can be attributed to age-related
constraints imposed by the motor and cognitive systems. Decreased motor performance is a
typical finding in studies on normal aging (Mattay et al., 2002; Perrot & Bertsch, 2007). Past
research has revealed that speech is affected as a result of aging (Jacewicz et al., 2009; Hoit &
Hixon, 1987; Searl et al., 2002). Old age typically causes systemic deterioration of the body

135
structures including the oral mechanism (Campbell, McComas, & Petito, 1973). Acoustic
analysis has confirmed that men over 70 years speak at a rate that is slower than that of younger
men (Ryan, 1972) and imprecise articulation is frequently implicated in the speech of elderly
individuals (Amerman & Parnell, 1990; Hartman & Danhauer, 1976). Sadagopan (2008)
compared the novel speech learning ability in younger vs. older age groups. A physiologic
measure through kinematic analysis (lip aperture coordination) and behavioural measures
(production accuracy and duration) were assessed on two consecutive days for 16 young and 16
elderly participants during the production of six novel non-words increasing in length and
complexity. Behaviourally, clear differences were noted between young and elderly participants
in the ability to accurately produce the longer, more complex non-words. Older speakers’
productions revealed a greater percentage of articulatory errors than young adults for four-
syllable non-words, suggesting that important age-related differences are present for repetition of
long, complex and, novel non-words. Elderly individuals also demonstrated longer durations for
non-word production than young adults, and this effect was more pronounced for longer, more
complex nonwords. Very few elderly individuals produced the requisite number of accurate
productions for kinematic analysis of the two most complex non-words, and were not subjected
to statistical analyses. The results of the current study are in close agreement with the findings of
Sadagopan.
Different models have been proposed to account for the cognitive-motor decline in elderly
individuals. The “information loss model” (Myerson et al., 1990) attributes the decreased motor
performance in the elderly population due to the loss of information at each of the various
information processing stages, thereby requiring more time to plan a motor response. The
“neural noise model” (Crossman & Szafran, 1956; Welford, 1985) explains the decreased motor
performance in elderly individuals based on the increased random activity in the brain due to
aging. Another model explaining the reduced motor output in elderly population is based on the
difference in attitudes and preferences shown by the elderly people in performing novel motor
tasks (Verhoff et al., 1984). Reduced working memory in older adults has been documented in
past studies (Kester et al., 2002; Jost et al., 2011). According to the inhibitory deficit hypothesis,
older individuals are unable to filter out task-irrelevant information from external sources, which,
in turn, reduces their memory capacity (Hasher & Zacks, 1988). Hasher and Zacks found that
older adults lack inhibitory control, thus allowing irrelevant information to enter their working

136
memory and distract them during memory tasks. In the present study, it is likely that all the
above models can be accounted for the decreased spatial learning (i.e., lower PPC) among the
elderly individuals.
The findings with respect to the practice conditions in the older group can be explained based
on similar reasons applicable for overall results of spatial learning. The advantage of constant
practice over blocked practice can be attributed to the role of repeated exposure to the
orthographic stimulus which could have aided in the retrieval of the phonological motor plan
pertaining to the speech phrase during the retention phrase. Thus, the continuous auditory and
visual representation of the target phrase offered by the constant practice could have facilitated
the participants to form an accurate cognitive representation of the phrase. The sequential
manner of practice in the blocked practice condition could have affected the participants in the
older group to a greater extent than the younger group in forming a mental imagery of the target
phrase. Past research has shown that younger adults have enhanced working memory in terms of
increased cognitive processing in comparison to older adults (Kahneman, 1973). This enhanced
cognitive processing may have made the younger participants less reliant on mental imagery and
hence, able to perform similarly across the four practice conditions.
The initial proposed hypothesis was “The PMLs that best facilitate spatial learning of a novel
speech utterance task will not be similar between a group of healthy younger individuals and a
group of healthy older individuals”. The findings of the spatial learning of the speech task
support this hypothesis. In the younger group, there was no difference in learning between the
participants in four practice conditions. In the older group, the participants in the constant
practice condition demonstrated better learning than participants in the blocked practice
condition.

Keyboard task - The mean PKC values of the younger age group during the retention phase
were significantly better that the older age group. The performance of younger, as well as older,
participants in the constant and variable practice conditions was better than participants in
random and blocked practice conditions. As there was no interaction effect between age of the
participants and practice condition, the performance of the younger and older participants across
the four practice conditions could not be analysed separately.
It is well known that aging is accompanied by impairments in sensorimotor (Ketcham &
Stelmach, 2001) as well as cognitive and perceptual functioning (e.g., Gunning-Dixon & Raz,

137
2000; Salthouse, 1985; Seidler, 2006; Perrot & Bertsch, 2007). For example, Perrot and Bertsch
investigated motor learning abilities of 31 younger (20–30 years) vs. 33 older adults (61-75
years) in a ball juggling task. The participants practised the juggling task for 12 sessions of 20
minutes each. Results revealed that the younger adults learned the task faster than the older
adults, and also older adults required more psychomotor ability to learn the juggling task in
comparison to the younger adults.
Older people perform complex motor tasks more slowly and less accurately than they once
did (Voelcker-Rehage, 2008). For example, Seidler (2006) examined young (18–31 years) and
old adults (65–80 years) in their ability to learn different joystick aiming tasks. Older adults
exhibited poorer performance and took longer to learn the visuomotor version of the joystick task
as compared to younger adults. Apparently, with increased difficulty level, age differences in
motor learning become more pronounced. The results of the present study are in agreement with
past studies indicating the effect of aging on spatial learning of non-speech tasks. The reasons
provided by the various models of aging could probably account for the decreased performance
of the older group in comparison to the younger group.
Fraser, Li, and Penhune (2009) assessed the retention performance of younger vs. older age
group on a multi-finger sequence task. Eighteen younger adults (M = 24 years) and 15 older
adults (M = 65 years) practised a sequence of keys on a piano keyboard through variable practice
in response to a pattern of visual stimulus that appeared on the computer monitor. The results in
terms of accuracy and reaction time revealed that older and younger adults demonstrated similar
performance during the retention test across days. The results of the present study also revealed
similar performance of both the age groups in variable practice condition; the current study also
found variable practice (along with constant practice) aided the learning of the keyboard, which
is in agreement with the findings of Fraser et al. It is likely that as the complexity of the practice
task increases, the need for the frequency of exposure to the visual and/or auditory
representations of the task also increases to form a mental imagery of the task. In the present
study, the practice conditions (i.e., constant and variable practice) which provided repeated
exposure to the auditory and visual representation of the task without any interruption of the
alternate tune could have facilitated in forming the mental imagery of the keyboard task.
The initial proposed hypothesis was “The PMLs that best facilitate spatial learning of a novel
musical keyboard entry (non-speech) task will not be similar between a group of healthy younger

138
individuals and a group of healthy older individuals”. The findings of the spatial learning of the
keyboard task do not support this hypothesis. The younger, as well as older, participants in the
constant and variable practice conditions demonstrated better learning than participants in
random and blocked practice conditions.

Temporal Learning
Speech task - The mean phi correlation values revealed that the temporal learning of the
younger group was better than the older group across the four practice conditions. There was no
difference in the performance of the young participants across the four practice conditions.
Similar to the spatial learning, it is likely that the increased cognitive resources of the
participants in the younger group could have facilitated temporal learning to a similar extent
across all the four practice conditions. The performance of the older participants in the constant
practice condition was better compared to the other three practice conditions. This can be
attributed to the role of mental imagery as in the case of spatial learning of the speech task. The
repeated visual and auditory representations of the speech phrase offered by the constant practice
condition could have helped the participants to develop a better cognitive representation of the
phrase, thereby facilitating the temporal learning. Another reason for the better performance of
the participants in the constant practice condition can be accounted by a timing model described
by Wing and Kristofferson (1973b). According to this model, there are two levels of timing: (1)
a central time keeper level and, (2) a motor implementation level. The centrally-generated
‘internal clock’ brings about the movement of desired goal duration by sending pulses via the
central nervous system. Wing (2002) mentioned that one of the factors affecting the internal
clock and motor implementation output is the increased working memory demands due to a
secondary task. It is possible that as the older participants in the constant practice condition had
to practice just one target phrase, the overall working memory demands would have been lesser
in comparison to the other three practice conditions which had interference from an alternate
tune.
The initial proposed hypothesis was “The PMLs that best facilitate temporal learning of a
novel speech utterance task will not be similar between a group of healthy younger individuals
and a group of healthy older individuals”. The findings of the temporal learning of the speech
task support this hypothesis. In the younger group, there was no difference in learning between
the participants in four practice conditions. On the other hand, in the older group, the participants

139
in the constant practice condition exhibited better learning than the participants in the other three
practice conditions.

Keyboard task - The mean phi correlation values revealed that there was no age effect or main
effect of practice condition. It is possible that temporal learning of the keyboard task could have
burdened the cognitive resources of the participants in both the age groups. This in turn could
have resulted in the keyboard task being equally difficult for participants in both age groups
across the four practice conditions. Two reasons for this finding are offered. First, there might
not be an age difference between the younger and older age groups in temporal learning of a
complex non-speech task. Second, the methodological limitations associated with calculating the
phi correlation approach could have possibly missed subtle aspects of temporal learning.
Previous research has measured relative and absolute durations of motor movements to
investigate temporal learning in motor tasks (Shea, Wulf, Park, & Gaunt, 2001; Shea, Lai,
Wright, Immink, & Black, 2001), it is likely that investigating temporal learning through such
outcome measures might be more revealing about temporal learning.
The initial proposed hypothesis was “The PMLs that best facilitate temporal learning of a
novel musical keyboard entry (non-speech) task will not be similar between a group of healthy
younger individuals and a group of healthy older individuals”. The current findings reveal that
there is no significant difference between the participants in four practice conditions across both
the groups, indicating that both the groups were similar in terms of temporal learning of the
keyboard task. This offers ground to reject this hypothesis.

Non-Clinical vs. Clinical group

The clinical group examined in the present study consisted of 16 individuals with PD. The
participants were randomly and equally assigned to the four practice conditions. The data
analysis was similar to the non-clinical group. The results of spatial and temporal learning in the
clinical group revealed no difference between the PD patients across the four practice conditions.
While it is possible that the groups performed similarly across practice conditions, a more likely
explanation is that the sample size was insufficient to detect small differences. Therefore, this
section focuses on a general comparison of the non-clinical and clinical group performances. The
results obtained through statistical comparison of the non-clinical and clinical groups are
exploratory in nature as unequal sample sizes were compared. In instances, where significant

140
differences were not obtained, both groups were compared using a descriptive statistical
approach. The similarities and/or differences between the non-clinical and clinical groups for
each of the major finding are discussed below.

Spatial Learning
Speech task - The mean PPC values of the constant, variable, random, and blocked practice
conditions in the clinical group were 78.1%, 72.5%, 74.8%, and 77.5%, respectively. The spatial
performance of the non-clinical group was significantly better than the clinical group. There was
no main effect of practice condition, suggesting that there were no significant differences
between the practice conditions across the clinical and non-clinical groups
A prominent characteristic of PD is hypokinetic dysarthria (Ho, Iansek, Marigliani, Bradshaw,
& Gates, 1998; Logemann, Fisher, Boshes, & Blonsky, 1978). The present group of participants
demonstrated hypokinetic dysarthria. The speech characteristics of hypokietic dysarthria include
hypophonia, monotonicity, breathiness/ hoarseness, imprecise articulation, and speaking rate
problems (Duffy, 2005). Connor, Abbs, Cole, and Gracco (1989) analysed the sequencing of
upper and lower lip and jaw peak velocities during the production of the nonword ‘sapapple’ and
reported decreased coordination of articulators in nine participants with hypokinetic dysarthria.
The past studies suggest that the hypokinetic dysarthria in patients with PD might affect the
speech output to varying extent depending on the severity (Duffy, 2005). Consistent with the
findings of the past studies examining the speech of individuals with hypokinetic dysarthria, the
results of the present study also indicate a decreased spatial performance of the participants with
hypokinetic dysarthria in comparison to the non-clinical group.
The initial proposed hypothesis was “The PMLs that best facilitate spatial learning of a novel
speech utterance task will be similar between a group of healthy individuals and a group of
individuals with hypokinetic dysarthria due to PD”. This hypothesis can be either supported or
rejected based on how the current findings are interpreted. One line of interpretation could be
that as there was no significant difference between the four practice conditions in clinical as well
as non-clinical groups, the PMLs equally facilitated the spatial learning of the novel speech
utterance in both the groups, thus supporting the hypothesis. The other line of interpretation
could be that none of the PMLs best facilitated the spatial learning of the speech utterance task,
as there was no difference between the four practice conditions in both the groups, thus rejecting
this hypothesis.

141
 Keyboard task - The mean PPC values of the constant, variable, random, and blocked practice
conditions in the clinical group were 82.2%, 77.8%, 54.2%, and 49.5%, respectively. The spatial
performance of the non-clinical group was significantly better than the clinical group. Studies
have shown that PD patients have difficulty in starting and executing movements (Wilson,
1925). In addition, rapid single-joint, simultaneous, and sequential movements are executed
abnormally in PD (Marsden, 1989; Solveri, Brown, Jahanshahi, & Marsden, 1992). For example,
Solveri et al. compared the motor learning abilities of 21 patients with PD and 23 age-matched
controls on a buttoning task. They found that both groups improved with practice, but the
performance of the control group was better than the PD group. The results of the present study
are consistent with the findings of Solveri et al. The physiologic constraints of the motor system
of the participants in the clinical group can possibly account for the decreased spatial
performance of the PD group.
The constant and variable practice conditions were significantly better that random and
blocked practice conditions across both the groups. As there was no significant interaction effect,
the differences between the practice conditions in non-clinical and clinical groups could not be
analysed separately. Sidaway, Gordon, Hopkins, Kershaw, Marean, & Wilkins (2006) compared
random vs. blocked practice conditions in four participants with PD. The participants practised
three 5-key press patterns on computer keyboard under both blocked and random practice
conditions. Retention tests after one day and one week revealed that superior performance was
exhibited by participants in the random practice condition. On the contrary, Lin, Sullivan, Wu,
Kantak, and Winstein (2007) found that participants with PD benefited from a blocked practice
in comparison to random practice on a lever movement task. Twenty adults with PD and 20 age-
matched adults practised three-lever movement tasks with either a blocked or a random practice
order. Retention tests revealed that participants in the control group who practised with a random
order performed more accurately than participants in the control group who practised with a
blocked order. However, for the PD group, the findings were reversed; participants who
practised with a blocked order performed more accurately than participants who practised with a
random order. The results of the current study reveal that there was no difference between the
random and blocked practice conditions.
The difference in findings between Lin et al. (2007) and Sidaway et al. (2006) can be
attributed to the ‘challenge point framework’ (CPF) (Guadagnoli & Lee, 2004). According to the

142
CPF, the practice conditions that facilitate the learning of a task depend on the difficulty of the
task. It is likely that the difference in difficulty of the tasks used in both the studies could have
resulted in different practice conditions facilitating the motor learning. The results of the present
study indicate that there was no significant difference between the random and blocked practice
conditions; this does not support the findings of Sidaway et al. or Lin et al. The small sample size
of the clinical group can be a reason for the lack of difference between the random and blocked
practice conditions in the present study. However, a comparison between random and blocked
practice conditions using a descriptive approach revealed that the performance of the random
practice group was slightly better than the blocked practice group, providing marginal support
for Sidaway et al. It could be possible that the computer keyboard press task used by Sidaway et
al. was similar to the musical keyboard task used in the present study leading to similar findings.
The initial proposed hypothesis was “The PMLs that best facilitate spatial learning of a novel
musical keyboard entry (non-speech) task will be similar between a group of healthy individuals
and a group of individuals with hypokinetic dysarthria due to PD”. Since the constant and
variable practice conditions were better than the random and blocked practice conditions in non-
clinical as well as clinical groups, the findings support this hypothesis.

Temporal learning
Speech task - The mean phi correlation values of the constant, variable, random, and blocked
practice conditions in the clinical group were 0.13, 0.10, 0.05, and 0.09, respectively. The
temporal performance of the non-clinical group was significantly better than the clinical group.
Basal ganglia disorders affect movement speed and rhythm of speech. (Kent & Rosenbek, 1982;
Netsell, 1983). Ludlow, Connor, and Bassich (1987) investigated the effects of two different
basal ganglia diseases on different aspects of speech timing (speech planning, initiation, and
production). Twelve patients with PD and 12 patients with Huntington’s disease (HD) were
compared with normal participants on four different speech timing tasks; reaction time, syllable
duration, sentence duration, and phrase duration. Results revealed that in PD as well as HD
patients, the control of sentence and phrase duration was impaired. The PD patients had
difficulties in altering sentence and phrase durations, but not syllables. On the other hand, HD
patients exhibited a global speech timing difficulty across sentences, phrases and syllables. The
researchers attributed the speech timing impairment in PD patients to the motor planning. The
results of the present study are consistent with the findings of Ludlow et al. The speech timing

143
deficits due to the impaired motor planning ability at the syllable and phrase levels could
possibly explain the decreased temporal learning in participants with PD in comparison to the
healthy participants. Despite the absent significant findings between the practice conditions, the
general trend of the results indicated that in clinical as well as non-clinical groups, the constant
practice condition facilitated spatial learning more than the other three practice conditions. In the
constant practice condition, the repeated practice of the target phrase could have strengthened the
formation of an internal template of a central clock as indicated by the timing model (Wing &
Kristofferson, 1973b). The low sample size, and the differences in severity of PD among
participants in the clinical group can be a probable reason for the lack of significant findings
between the practice conditions.
The initial proposed hypothesis was “The PMLs that best facilitate temporal learning of a
novel speech utterance task will be similar between a group of healthy individuals and a group
of individuals with hypokinetic dysarthria due to PD”. Again this hypothesis can either be
rejected or accepted based on the direction of the interpretation of the findings. The lack of
significant difference between the four practice conditions in clinical as well as non-clinical
groups could mean that all the PMLs were equally facilitative in temporal learning of the speech
task, which would support of the hypothesis. Alternatively, this finding could also be interpreted
that none of the PMLs best facilitated the temporal learning of the speech task, thus rejecting the
hypothesis.

Keyboard task - The mean phi correlation values of the constant, variable, random, and
blocked practice conditions in the clinical group were 0.16, 0.24, 0.10, and 0.15, respectively.
There was no significant difference in the temporal performance between non-clinical group and
clinical group and also there was no main effect of practice condition. Even though there were no
significant difference between the practice conditions, the general trend of the results indicate
that the best performance for the clinical participants was found for those undertaking variable
practice, this was also the case for the non-clinical group which showed the best performance for
variable, as well as constant practice. As in the case of temporal learning of the speech phrase,
the low sample size, and the differences in severity of PD among participants in the clinical
group could have contributed to the lack of significant findings between the practice conditions.
The initial proposed hypothesis was “The PMLs that best facilitate temporal learning of a
novel musical keyboard entry (non-speech) task will be similar between a group of healthy

144
individuals and a group of individuals with hypokinetic dysarthria due to PD”. The current
findings reveal that there is no significant difference between the participants in four practice
conditions across both the groups, and the descriptive comparison between the clinical and non-
clinical groups indicated that both the groups were similar in terms of temporal learning of the
keyboard task. Thus, the findings support this hypothesis.

NASA task load index

There were no specific hypotheses with regards to NASA task load index. However, a part of
the current research focused on investigating the perceived difficulty of the speech and non-
speech-motor learning tasks by the participants in relation to their retention performance by
using NASA task load index.

Speech task - Based on a median split of the index scores according to the age of the
participants, the results revealed that the mean index score of the older group was higher than the
older group, which suggests that the older group perceived the task to be more difficult than the
younger group. There was also a main effect of the practice condition, with the index scores of
the random and blocked practice conditions being significantly higher than the constant practice
condition. As there was no significant interaction effect, the differences in the index scores
between the four practice conditions could not be analysed separately in younger and older age
groups.
The decreased motor and cognitive resources as a result of aging could have resulted in the
older group perceiving the speech task to be more difficult than the younger group. In regards to
the practice condition, the repeated auditory and visual representations provided by the constant
practice condition could probably account for the reason that it was perceived to be the least
complex practice condition. In the case of random and blocked practice conditions, the additional
task load of practising a second alternate phrase along with the target phrase could have made the
participants in both the age groups perceive these two practice conditions as being the most
difficult. Even though the variable practice condition was not significantly lower than the
random and blocked practice condition, visual inspection of the data revealed that the index
scores of the variable practice condition was lower than random and blocked practice condition
suggesting that it may have been perceived to be less complex than random and blocked practice
conditions.

145
 Keyboard task - The results revealed that there was no age effect. The lack of experience in
playing a musical keyboard could have led to the task being equally difficult by the participants
in younger as well as older age groups. However, there was a main effect of practice condition,
with the index scores of the random and blocked practice conditions being higher than the
constant practice condition. As in the case of speech task, the additional task load of practising
an alternate phrase along with the target phrase could have made the participants in both the age
groups perceive random and blocked practice conditions as being the most difficult.

Speech vs. keyboard tasks - The comparison of the index scores between the speech and
keyboard tasks indicated that the keyboard task was considered to be significantly more difficult
than the speech task. Both the tasks offered auditory and visual representation during the practise
regime. However, a major difference among these two tasks was that the speech task was
linguistic in nature as it offered an orthographic representation of the target phrase. This could
have helped the participants to visualize the speech task more easily than the keyboard task. The
limited ‘imageability effect’ offered by the keyboard task could account for it to be perceived
more difficult that the speech task. Also, since speech is a highly practised task in comparison to
the keyboard task, the participants could have felt the speech task to be easier than the keyboard
task.

The non-speech vs. speech debate

There is ongoing debate regarding the clinical basis and utility for the use of non-speech-
oromotor tasks to assess and treat motor-speech disorders (MSDs) (McCauley, Strand, Lof,
Schooling, & Frymark, 2009). One line of thought argues against the use of non-speech-
oromotor tasks in assessment and treatment of MSDs (Forrest, 2002; Clark, 2003; Ziegler, 2003;
Weismer, 2006; Lass & Pannbacker, 2008; Ruscello, 2008; Powell, 2008). According to this
view, the acoustic signal is an integral part of speech-motor control. Non-speech-oromotor tasks
do not involve speech production so it is unlikely that these tasks provide insight to the speech
productions deficits in MSDs.
The other line of thought advocates the use of non-speech-oromotor tasks in assessment of
MSDs (Folkins, 1985; Folkins, Moon, Luschei, Robin, Tye-Murray, & Moll, 1995; Ballard,
Robin, & Folkins, 2003). According to this view, the underlying cause of MSDs is a motor
problem. As such, the inclusion of speech tasks in the assessment may result in a failure to

146
separate motor from linguistic factors in speech performance. Therefore, use of speech tasks
exclusively to evaluate speech-motor control may not reflect the underlying motor control
problem. One recurring issue in this debate is the need to evaluate parts of the speech mechanism
independently of other parts. Hence, inclusion of linguistic factors in an evaluation task may
impede the understanding of a suspected motor control deficit.
Ziegler (2003) opposed using non-speech tasks in the assessment and treatment of MSDs. He
proposed a task dependent model of speech-motor control, whereby movements of the tongue,
lips, and larynx are controlled in fundamentally different ways depending on the particular motor
activity. Furthermore, the task dependent model explains that the various subsystems of speech
production (respiratory, phonatory, resonatory and articulatory subsystems) are separate to the
extent that each of them has unique properties, are subserved by a neural circuitry, and can be
impaired selectively after brain lesions. Weismer (2006) also supports Ziegler’s view on task
specificity. He suggests that there is neither theoretical nor clinical support for implementation of
non-speech-oromotor tasks in assessment and treatment of MSDs. Weismer further states that the
relation between disordered speech and speech acoustics cannot be observed in studies of non-
speech-oromotor behaviour, but rather in studies of speech production in persons with MSDs.
Thus, he suggests that the underlying speech deficits in MSDS are best assessed and treated
using speech tasks rather than non-speech based tasks.
The concept of task specificity is also strongly advocated by Bunton (2008). She explained
the differences between the speech mechanism and non-speech mechanism based on four
perspectives; (1) movement characteristics of non-speech oral motor behaviors and speech
production, (2) treatment studies, (3) basis of motor learning, and (4) neuroanatomical
underpinnings. Based on the data from these domains, she suggested that there is little theoretical
or clinical evidence to recommend non-speech activities in the practice of Speech-Language
Pathology.
The application of evidence based practice concerning the use of non-speech-oromotor
treatment was studied by Lass and Pannbacker (2008). They conducted a systematic literature
search using the electronic databases and reviewed a total of 45 articles/reports that were
published between 1981 and 2006 in peer-reviewed and non-peer-reviewed journals. They
concluded that evidence is either weak or lacking for the use of non-speech tasks in the treatment

147
of speech disorders. Similar conclusions were reached by Ruscello (2008), Powell (2008) and
Wilson, Green, Yunusova, and Moore (2008).
All the above studies are suggestive of task specificity of speech. However, the alternative
line of thought promotes the use of non-speech-oromotor tasks in assessment of speech
disorders. Ballard, Robin, and Folkins (2003) proposed an integrative model of speech- motor
control. According to this model, speech and volitional non-speech-motor control are integrated
into the functioning of a more general motor system where neural and behavioural systems
demonstrate areas of overlap. Folkins (1985) postulated an integrated motor approach to speech
production in which speech is organized ultimately to produce the holistic behaviour of
communication. Folkins’ model was developed to argue against the need to use linguistic units
as organizing structures for the motor aspects of speech. The integrative model does not claim
complete task-dependence or task-independence, rather it takes a stand between the two.
According to this model, certain volitional non-speech tasks share principles in common with
speech and therefore with speech-motor anomalies (e.g., dysarthria). At complex behavioural
levels, there must be overlapping functional components and, therefore, overlapping and
integrative neural pathways or networks. Folkins et al. (1995) suggested that in order to assess
motoric deficits in an individual with MSD it is necessary to separate the motoric deficits from
the psycholinguistic deficits if present. Non-speech tasks can be designed to measure the pure
motoric deficits and give better insight to understanding the nature of the prevailing MSD.
Netsell (1986) also stressed the importance of using non-speech-oromotor tasks as valid
assessment tools in individuals with MSDs. Netsell highlighted the potential benefits of non-
speech tasks in differential diagnosis and as specific disease markers in order to find out the
underlying neuropathophysiology of the speech-motor system. More recently, McCauley, Strand,
Lof, Schooling and Frymark (2009) examined the peer-reviewed literature from 1960 to 2007 for
articles on the use of non-speech-oromotor exercises (NSOMEs) that affect speech physiology,
production, or functional outcomes (i.e., intelligibility). They found insufficient evidence to
support or refute the use of NSOMEs to assist with improving speech-motor control.

The non-speech vs. speech debate in the context of the present study
Lof and Watson (2008) conducted a survey to investigate the usage of NSOMSEs among
speech language pathologists (SLPs) in USA. They found that 85% of the SLPs who responded
to the survey used NSOMEs to treat speech disorders in children. In another similar survey,

148
Mackenzie, Muir and Allen (2010) found that 86% of the SLPs in UK who responded to the
survey used NSOMEs to treat speech problems related to dysarthria. As at present, there is
insufficient evidence to support or refute the use of NSOMEs; SLPs continue to use NSOMSEs
to treat a wide facet of speech disorders in children as well as in adults. NSOMEs encompass a
wide range of activities targeted to improve muscle strength and coordination of oral structures.
Hodge (2002) indicated that NSOMEs are a collection of stimulation techniques and procedures
that are designed to influence the resting posture and/or movement of the lips, jaw, and tongue.
An ideal way to resolve this ongoing debate would be to train participants using non-speech-
based tasks and observe for transfer in the speech tasks. The current research was designed to
assess retention benefits of selected PMLs rather than transfer benefits. However, the present
study found that constant practice was beneficial for spatial and temporal learning in speech as
well as non-speech tasks. So it may be that if certain PMLs are followed, one might expect
transfer benefits from the non-speech to the speech domain. Admittedly, the non-speech task
used in this study was not an oro-motor task; still it seems that some non-speech as well as
speech tasks respond consistently to specific PMLs. As most SLPs continue to use NSOMEs of
varying complexity to treat speech disorders, it would be worthwhile to explore the use of
constant practice condition in implementing such complex non-speech based activities in the
light of the present study. However, caution should be exercised in generalizing the results of the
current study to NSOMSEs, as the present study used a limb-based task as the non-speech task
instead of the usual oral-based non-speech task. Further research along the lines of the present
study using an oral-based non-speech task instead of a limb-based task can contribute further
evidence to support or refute the ongoing debate.

Limitations of the Study

1. One of the main limitations in the current study seems to be with regards to the approach
adopted to estimate temporal learning. Temporal learning can be estimated by at least four
different ways. First, through verbal estimation of the stimulus duration (e.g., verbal
estimation of the duration of a tone or empty intervals). Second, through temporal
discrimination tests (e.g., presenting two tones and determining whether the second tone is
shorter than the first tone). The third way is through temporal production (e.g., subject is
asked to produce a certain interval by pressing a button), and (4) temporal reproduction,

149
 (e.g., subject is presented with a stimulus of certain duration and is asked to reproduce
that duration) (Salman, 2002). In the current study, the temporal learning was evaluated
through temporal reproduction, as this method is frequently used in the motor learning
research (e.g., Wulf, Lee & Schmidt, 1994). It is likely that estimating temporal learning
through any of the other three methods would have provided additional information. For
example, as none of the participants had any experience in playing the keyboard,
assessing temporal learning by having the participants to verbally estimate the duration of
the keyboard tune could have resulted more accurate results than temporal reproduction of
the keyboard tune.
2. Another limitation pertains to the use of phi correlation as a measure of temporal learning.
Other measures of temporal learning, like relative and absolute duration, have been
frequently reported in the motor learning literature (Lai, Shea, Wulf, & Wright, 2000;
Wulf, Lee & Schmidt, 1994; Adams & Page, 2000). Adams and Page (2000) used
absolute duration to compare constant vs. variable practice conditions in learning the
utterance “Buy Bobby a poppy” with a specific overall utterance duration. Results
revealed that the group undergoing constant practice were less successful in their training
compared to the variable practice group. In the present study, absolute duration measures,
like speech segment duration and pause durations, may have provided additional details
regarding temporal learning. Also, to determine the synchrony between the participants’
production of the target phrase/and the original examples of the target phrase/tune, the
alignment of the participant productions and the original example of the target phrase/tune
occurred at the onset point of the acoustic waveform. However, it could have been
possible that the final segment of the participant productions could have been more
synchronous to the target tune/phrase than the initial segment. In this case, aligning the
participant productions and the target phrase/tune at the offset point of the acoustic
waveform may have resulted in higher phi correlations.
3. The third limitation is with regards to the length of the training (i.e., acquisition phase). In
the present study, retention was assessed following two consecutive days of training.
Some studies that have investigated speech-motor learning assessed retention following
treatment/training ranging from 2 days to one week (Adams & Page, 2000; Pendt, Reuter,
& Muller, 2011; Rostami & Ashayeri, 2009). For example, Pendt et al. (2011) compared

150
 the timing release abilities of 19 patients with Parkinson's disease and 19 healthy control
group participants on a throwing task. The participants performed 200 throws per day
through blocked practice and the training lasted for a period of five days. Retention test
results after seven months revealed accurate temporal learning of the throwing task by the
participants in both the groups. In the present study, additional training of the participants
(e.g., three days) may have resulted in improved/increased temporal learning of the speech
and non-speech tasks.
4. The fourth limitation was that the sample size of the PD group was small. A low sample
size renders low statistical power which limits the ability to confidently reject the null
hypothesis (Ellis, 2010). This is a possible reason for no significant differences being
found between the practice conditions in the clinical group. Notwithstanding, the current
study indicates that if there were differences between the conditions, they are small.
5. The importance of pre-practice in motor learning has been documented by previous
studies (Edwin, Karyll, Lise, & Gary, 1981; Murray, McCabe, & Ballard, 2011; Bricker-
Katz, McCabe, Lincoln, & Ballard, 2011. Even though pre-practice instructions were
provided to participants prior to the start of the experiment, it was not done rigorously and
this could have been another limitation of the study.
6. A common method of splitting the data into two groups is by performing a median split.
In some occasions, a quartile split is performed, wherein the data is split into four groups
such that 25% of the observations are in each group (Altman & Bland, 1994). In the
current study, a median split was performed to divide the participants into two age groups,
and this could have been one of the limitations of the study. It is likely that performing a
quartile split instead of the median split could have placed the participants into four
different age groups, thus offering more specific information on the spatial and temporal
learning abilities of each age group.
7. The complexity of the tasks involved in the practice could have possibly influenced the
effects of each practice condition. For example, participants in the constant practice
condition had to practise one spatial pattern and one temporal pattern. Whereas,
participants in the random practice condition had to practice two spatial patterns and two
temporal patterns in a random manner. It is likely that the easier nature of the task in the

151
 constant practice condition could have led to better learning outcomes in comparison to
the other practice conditions.
8. The current study infers the learning outcomes of the participants in each of the four
practice conditions based on the data reported at the end of the retention phase. However,
there were no data reported with reference to skill mastery of the participants at the end of
the acquisition phase. This could be a potential limitation because the amount of retention
(or learning) of a motor skill is typically determined in reference to the skill acquisition.
Without baseline data at the end of the acquisition phase, quantification of the amount of
motor learning demonstrated at the end of the retention phase is likely to be less precise.
9. The final limitation is that patients with PD were selected as the clinical group.
Parkinson’s disease (PD) is a neurodegenerative disorder caused due to the dopamine
deficiency in the substantia nigra (Duffy, 2005). This gradually affects the brain’s ability
to generate body movements. It may be difficult to establish the efficacy of PMLs in this
clinical cohort, as the retention benefits of PML might disappear over the course of time
due to neurodegeneration. Using PMLs to train/re-train motor tasks in a different clinical
cohort like patients with stroke might provide more information about the efficacy of
PMLs. Patients with stroke present with an impaired but stable motor system, as stroke is
not a neurodegenerative condition there are more chances for the patients with stroke to
retain the learned motor skills.

Directions for future research

The present research can be extended in many directions. First, the current study compared
four practice conditions in learning a speech and a non-speech task. There are other PMLs with
respect to attentional focus (internal vs. complex), target complexity (simple vs. complex),
practice amount (small vs. large) (Mass et al., 2008) which were not addressed in the current
study. It is likely that each PML may contribute to motor learning in a unique manner (Mass et
al., 2008). For example, Freedman, Mass, Caligiuri, Wulf, and Robin (2007) compared the use of
external vs. internal focus of attention in learning oral and limb based motor tasks. Two groups
of 23 participants were administered hand and tongue impulse force control tasks. Each group
was randomly assigned to either an internal or an external focus of attention. Participants were
required to exert rapid pressure bursts to achieve a target force level of 20% of their maximal
strength. Results revealed an advantage of an external focus over internal focus of attention for

152
both the hand and tongue control tasks. The present study can be extended further by
investigating the role of each of these principles in speech and non-speech-motor learning tasks.
This line of future research might help to determine the PML which best facilitates speech-motor
learning and thus can be incorporated in developing speech therapy protocols.
Second, the present study included adult participants in clinical and non-clinical groups.
However, the role of PMLs in learning complex speech tasks in children has still to be
investigated. Mass and Farinella (2012) treated children with AOS using random and blocked
practice conditions. The researchers found contrasting results in comparison to Knock et al.
(2000) who recruited adults with AOS. It is plausible that the effects of PML in learning
complex speech tasks are different for children than for adults. Children with speech impairment
represent a substantial percentage among the school age children (McLeod, Harrison, McAllister
& McCormack, 2007). McLeod et al. analysed 4,983 parental reports and 3,276 teacher reports
and found that 25.2% of the children aged between four and five years had some sort of
expressive speech and language impairments. In the USA, almost 91% of SLPs in schools
indicated that they saw pupils with phonological/articulation disorders (ASHA, 2006). Future
research investigating the beneficial role of PMLs in speech-motor learning among children may
prove useful in rehabilitating children with speech impairment.
Third, the speech task in the present study involved learning a complex meaningless speech
phrase. According to CPF, the learning outcome of a task is highly dependent on the task
complexity (Lee & Guadagnoli, 2004). Adams and Page (2000) compared constant vs. variable
practice conditions in learning a meaningful phrase “Buy Bobby a poppy” with a specific overall
utterance duration. They found that the group undergoing constant practise were less successful
in their training compared to the variable practise group. The current study used a meaningless
phrase to separate the effect of linguistic familiarity on speech-motor learning and found
constant practise to facilitate speech-motor learning. It is likely that extending the current study
by including a meaningful speech phrase as a practice stimulus might result in other practise
conditions favouring speech-motor learning. Results obtained by comparing practice conditions
in learning a meaningful phrase may be more representative of the speech-motor learning rather
than using a non-meaningful phrase.
Fourth, further studies could include three data collection points during the motor learning
process, (1) at the beginning of the practice sessions, (2) at the end of the practice sessions, and

153
(3) during the retention phase. This arrangement could prove useful in determining the influence
of the PMLs on motor learning by tracking the participants’ motor learning performance from
the beginning of the practice session until the end of the retention phase.
Fifth, an important concept, which needs to be investigated, is the concept of response
generalization to other motor tasks. In the current study, based on the retention data alone, it was
determined that the constant practice condition was beneficial in comparison to the other practice
conditions. However, it is possible that the other practice conditions, which were not responsive
to the particular motor tasks in the current study, could be beneficial in learning simpler motor
tasks, which are used more frequently in everyday life.
Finally, an important concept of motor learning which remains to be investigated is the
transfer effect. A possible way to address the ongoing non-speech vs. speech debate would be to
investigate the transfer effect of non-speech tasks across speech tasks. Caviness, Liss, Adler, and
Evidente (2006) studied the task specificity of speech to address the ongoing non-speech vs.
speech debate. They compared speech and non-speech tasks in healthy controls and in
individuals with PD through a measure known as Electroencephalographic-Electromyographic
(EEG-EMG) coherence. Coherence is based on a measure of linear relatedness between two
waveforms as a function of frequency. This measure is thought to reflect coupling between
neural electrophysiological mechanisms in the control of non-speech and speech movement
production. They recruited 20 healthy participants and 20 individuals with PD for the study, all
the participants were required to carry out two non-speech and four speech tasks. During the
non-speech and speech production tasks, the EEG-EMG coherence was simultaneously
measured. They found varied coherence values within both the speech and the non-speech tasks
in both the groups, which supported the notion of task specificity of speech. However, the
researchers did not address the issue of transfer. Future research could be designed to optimally
train non-speech movements using PMLs and then observe transfer to comparable speech tasks
through EEG-EMG coherence. A study of this nature should help to address “the ongoing non-
speech vs. speech debate”.

154
Clinical Implications

Based on the findings of the current study, it is likely that a constant practice regime could be
possibly beneficial in learning complex and novel speech motor tasks. Secondly, the speech
learning tasks incorporating certain PMLs could lead to different learning outcomes in younger
vs. older participants. Finally, the results suggest that it might be easy for patients/participants to
learn spatial and temporal aspects individually rather than having to learn both the aspects
simultaneously due to possible spatial-temporal trade-off.
As the current study is translational in nature, caution should be exercised in generalizing the
above findings to clinical situations. It is likely that the above practice conditions could interact
with factors like amount of practice/treatment, length of practice, nature of speech disorders, age
of the patients undergoing treatment, and severity of disorder. More research is required along
these lines before generalizing the above findings to clinical conditions.

Conclusion
The summary of the findings are: First, in terms of spatial learning, the speech task was
learned better than the keyboard task. Second, in general, participants in the constant practice
condition learned the speech as well as non-speech tasks better than the participants in the other
three practice conditions. Third, there was a spatial-temporal trade-off as indicated by low phi
correlation scores in speech as well as non-speech tasks (i.e. the temporal learning was
compromised in comparison to the spatial learning). Fourth, there was an age effect, with the
motor learning outcomes being better in the younger age group than in the older age group, and
finally, there were no apparent differences in the effects of PMLs on speech and non-speech
motor learning between non-clinical and clinical groups.
Stages of motor learning
Recall, motor learning is a continuous process. An individual trying to learn a novel motor
skill gradually progresses through three different stages of motor learning: the cognitive stage,
the associative stage, and the autonomous stage. A logical question to be posed at the end of the
current study is: in what stage of motor learning can the participants be placed with respect to
spatial and temporal learning? An attempt is made to place the participants in one of the three
stages of motor learning based on the retention test performance on the spatial and temporal
aspects of the speech and keyboard tasks.

155
 Spatial learning - In case of speech as well as keyboard tasks, the participants were aware of
their goals, demonstrated consistent performance and, were able to detect errors during their
performance. Some of the participants were even able to execute the tasks automatically to
certain extent without much conscious effort. More than half of the participants were not
dependent on the feedback to perform the speech and keyboard tasks. The participants had a
success rate of 50-60%. Based on the movement characteristics exhibited by the participants, it is
likely that the participants were in the associative stage of motor learning with respect to the
spatial learning.

Temporal learning - In speech as well keyboard tasks, the participants were not aware of their
goals. Some of the participants did not even attempt to temporally match their productions with
the target phrase/tune. The participants exhibited highly inconsistent performance during the
production trials, and were not able to detect their errors. It seemed that the participants were still
heavily dependent on feedback to temporally align their productions with the target phrase/tune
during the production trials. On an overall note, the participants had a success rate of 15-25%.
Based on the movement characteristics exhibited by the participants, it is likely that the
participants were in the cognitive stage of motor learning with respect to the temporal learning.

156
References

Adams, J. A. (1971). A closed-loop theory of motor learning. Journal of Motor Behaviour,

3,111-150.
Adams, S. G., & Page, A. D. (2000). Effects of selected practice and feedback variables on
speech motor learning. Journal of Medical Speech-Language Pathology, 8(4), 215–220.
Adams, S. G., Page, A. D., & Jog, M. (2002). Summary feedback schedules and speech motor
learning in Parkinson’s disease. Journal of Medical Speech-Language Pathology, 10(4),
215–220.
Altman, D. G., & Bland, J. M. (1994). Statistics Notes: Quartiles, quintiles, centiles, and other
quantiles. BMJ, 309, 996.
Amerman, J. D., & Parnell, M. M. (1 990). Clinical impressions of the speech of normal elderly
adults. British Journal of Disorders of Communication, 25, 34-43.
Annett, J. (1969). Learning a pressure under conditions of immediate and delayed knowledge of
results. Quarterly Journal of Experimental Psychology, 11, 3-15.
Anshel, M. H. (1978). Effect of aging on acquisition and short-term retention of a motor skill.
Journal of Perceptual and Motor Skills, 47, 993-994.
ASHA. (2006). 2006 Schools Survey report: Caseload characteristics. Rockville, MD: ASHA.
ASHA-Ad Hoc committee on Apraxia of Speech in Children. (2007). Childhood Apraxia of
Speech-Technical Report. Rockville, MD: ASHA.
Baddeley, A. D., & Longman, D. J. A. (1978). The influence of length and frequency of training
session on the rate of learning to type. Ergonomics, 21, 627–635.
Baker, E. (2012a). Optimal intervention intensity. International Journal of Speech-Language
Pathology, 14, 401–409.
Baker, E. (2012b). Optimal intervention intensity in speech-language pathology: Discoveries,
challenges, and unchartered territories, International Journal of Speech-Language
Pathology, 14, 478–485

Ballard, K. J., Maas, E., & Robin,D. A. (2007). Treating control of voicing in apraxia of speech
with variable practice. Aphasiology, 21(12), 1195–1217.
Ballard, K. J., Robin, D. A., Woodworth, G., & Zimba, L. D. (2001). Age-related changes in
motor control during articulator visuomotor tracking. Journal of Speech, Language, and
Hearing Research, 44, 763–777.
Ballard, K. J., Robin, D.A., & Folkins, J. W. (2003). An integrative model of speech motor
control: A response to Ziegler. Aphasiology, 17, 37-48.

157
Barnette, J. J. (2006). [PDF document]. Effect Size and Measures of Association. Retrieved
http://www.eval.org/summerinstitute/06SIHandouts/SI06.Barnette.TR2.Online.pdf
Baron, A., & Menich, S. R. (1985). Reaction times of younger and older men: Effects of
performances. Journal of the Experimental Analysis of Behaviour, 44, 1-14.
Barton, J. W. (1921). Smaller versus larger units in learning the maze. Journal of Experimental
Psychology. 4, 414-424.
Battig, W. F. (1972). Intratask interference as a source of facilitation in transfer and retention. In
R.F. Thompson & J.F. Voss (Eds.), Topics in learning and performance (pp. 131-159).
New York: Academic Press.
Bilodeau, E. A., Bilodeau, I. M., & Schumsky, D. A. (1959). Some effects of introducing and
withdrawing knowledge of results early and late in practice. Journal of Experimental
Psychology, 58, 142-144.
Bislick, L. P., Weir, P. C., Spencer, K., Kendall, D., & Yorkston, K. M. (2012). Do principles of
motor learning enhance retention and transfer of speech skills? A systematic review.
Aphasiology, 26(5), 709-728.
Boss, G. R., & Seegmiller, J. E. (1981). Age-Related Physiological Changes and their Clinical
Significance. The Western Journal of Medicine, 135(6), 434–440.
Boyce, B. A. (1991). The effects of an instructional strategy with two schedules of augmented
KP feedback upon skill acquisition of a selected shooting task. Journal of Teaching in
Physical Education, 11, 47-58.
Brady, F. (2008). The contextual interference effect and sport skills. Perceptual and Motor
Skills, 106, 461-472.
Brashers-Krug, T., Shadmehr, R., & Bizzi, E. (1996). Consolidation in human motor memory.
Nature, 382, 252–255.
Breitenstein, C., Daum, I., & Schugens, M. M. (1996). Altersunterschiede beim motorischen
Lernen. Zeitschrift fur Gerontopsychologie und-psychiatrie, 1, 33- 41.
Breslin, G., Hodges, N. J., Steenson, A., William, A. M. (2012). Constant or variable practice:
Recreating the especial skill effect. Acta Psychologica, 140, 154–157.
Bricker-Katz, G., McCabe, P., Lincoln, M., & Ballard, K. (2011). Principles of motor learning in
stuttering treatments. In B. Maassen & P. van Lieshout (Eds.). Proceedings of the 6th
International Conference on Speech Motor Control, 17, 55.
Briseno, M. P., Diaz, R., Romo, C. A., & Ruiz, F. J. (2010). Sex-related differences in motor
learning and performance. Behaviour and Brain Function, 6(1), 74.
Brody, H. (1955). Organization of the cerebral cortex. III: A study of aging in the human
cerebral cortex. The Journal of Comparative Neurology, 102, 511–556.

158
Bromage, B. K., & Mayer, R. E. (1986). Quantitative and qualitative effects of repetition on
learning from technical text. Journal of Educational Psychology, 78, 271-278.
Brown, J. D. (2008). Effect size and eta squared. Shiken: JALT Testing & Evaluation SIG
Newsletter, 12(2), 38-43.
Bunton, K. (2008). Speech versus Non-speech: Different Tasks, Different Neural Organization.
Seminars in Speech and Language, 29(4), 267–275.
Burdick, K. J. (1977). Effects of massed and distributed practice on the learning and retention of
a novel gross motor skill. Master’s Thesis, Western Illinois University.
Calhoun, K. H., Gibson, B., Hartley, L., Minton, J., & Hokanson, J. A. (1992). Age-related
changes in oral sensation. The Laryngoscope, 102(2), 109–116.

Campbell, M. J., McComas, A. J., & Petito, F. (1973). Physiological changes in ageing muscles.
Journal of Neurology, Neurosurgery, and Psychiatry, 36, 174-182.
Caviness, J. N., Liss, J. M., Adler, C., Evidente, V. (2006). Analysis of high-frequency
electroencephalographic- electromyographic coherence elicited by speech and oral non-
speech tasks in Parkinson’s disease. Journal of Speech, Language and Hearing Research,
49, 424-438.
Clark, H. M. (2003). Neuromuscular treatments for speech and swallowing: A tutorial. American
Journal of Speech-Language Pathology, 12, 400–415.
Čoh, M., Jovanović-Golubović, M., & Bratić, M. (2004). Motor learning in sport. Physical
Education and Sport, 2(1), 45 – 59.
Coker, P. C. (2009). The effects of an experiential learning program on the clinical reasoning
and critical thinking skills of occupational therapy students. Retrieved from ProQuest
Dissertations and Theses database.
Connor, N. P., Abbs, J. H., Cole, K. J., & Gracco, V. L. (1989). Parkinsonian deficits in serial
multiarticulate movements for speech. Brain, 112, 997-1009.
Vidal, C. J. L., Teulings, H. L., & Stelmach, G. E. (1998). Elderly subjects are impaired in
spatial coordination in fine motor control. Acta Psychologica, 100, 25–35.
Cook, A., Brauer, S., & Woollacott, M. (2000). Predicting the probability for falls in community-
dwelling older adults using the Timed up & go Test. Physical Therapy, 80(9), 896-903
Crossman, E. R. F. W., & Szafran, J. (1956). Changes with age in the speed of information
intake and discrimination. Experientia Supplementum, 4, 128-135.

Darley, F. L., Aronson, A. E., & Brown, J. R. (1975). Motor speech disorders. Philadelphia: W.
B. Saunders.

Davis, B. et al. (2000). The Nature and classification of skill. In B. Davis et al. (Ed.), Physical
Education and the study of sport. (4th ed. pp. 284-285). Spain: Harcourt Publishers.

159
Dean, N. J., Kovacs, A. J., Shea, C. H. (2008). Transfer of movement sequences: Bigger is
better. Acta Psychologia, 127, 355-368.
Dick, M. B., Beth, R. E., Shankle, R. W., Dick-Muehlke, C., Cotman, C., & Kean, M.-L. (1996).
Acquisition and long-term retention of a gross motor skill in Alzheimer’s disease patients
under constant and varied practice conditions. Journal of Gerontology, 51, 103– 111.
Dickman, S. J., & Meyer, D. E. (1988). Impulsivity and speed-accuracy tradeoffs in information
processing. Journal of Personality and Social Psychology, 54, 274–290.
Dollaghan, C., & Campbell, T. F. (1998). Nonword repetition and child language impairment.
Journal of Speech, Language, and Hearing Research, 41, 1136–1146.
Driskell, J. E., Cooper, C., & Moran, A. (1992). Effect of overlearning on retention. Journal of
Applied Psychology, 77, 615-622.

Duffy, J.R. (2005). Motor Speech Disorders: Substrates, Differential Diagnosis, and
Management. (2nd Ed). St Louis, MO: Mosby.

Edeal, D. M., & Gildersleeve-Neumann, C. E. (2011). The importance of production frequency

in therapy for childhood apraxia of speech. American Journal of Speech-Language
Pathology, 20, 95–110.
Edwin, A. L., Karyll, N. S., Lise, M. S., & Gary, P. L. (1981). Psychological Bulletin 90(1), 125-
152.
Ellis, P. D. (2010). The Essential Guide to Effect Sizes: An Introduction to Statistical Power,
Meta-Analysis and the Interpretation of Research Results. Cambridge: Cambridge
University Press.
Emanuel, M., Jarus, T., & Bart, O. (2008). Effect of focus of attention and age on motor
acquisition, retention, and transfer: A randomized trial. Physical Therapy, 88(2), 251-
260.
Enderby, P. (1983). Frenchay Dysarthria Assessment. San Diego, CA: College-Hill Press.
Enderby, P. (2012). How much therapy is enough? The impossible question! International
Journal of Speech-Language Pathology, 14, 432–437.
Epstein, W. (1962). A further study of the influence of syntactical structure on learning. The
American Journal of Psychology, 75(1), 121-126.
Ericsson, K. A., Krampe, R. T., & Tesch-Romer, C. (1993). The Role of Deliberate Practice in the
Acquisition of Expert Performance. Psychological Review, 100(3), 363-406.

Fairbrother, J. T. (2010). Fundamentals of motor behaviour. Champaign, IL: Human Kinetics.

Falduto, L. L., & Baron, A. (1986). Age-related effects of practice and task complexity on card
sorting. Journal of Gerontology, 41, 659-661.

160
Field, A. (2009). Discovering Statistics using SPSS ( 3rd ed.). London, UK: SAGE Publications
Ltd.
Fitts, P. M. (1954). The information capacity of the human motor system in controlling the
amplitude of movement. Journal of Experimental Psychology, 47, 381–391.
Fitts, P. M., & Posner, M. I. (1967). Human performance. Belmont, CA: Brooks/Cole.
Folkins, J. W. (1985). Issues in speech motor control and their relation to the speech of
individuals with cleft palate. Cleft Palate Journal, 22, 106-122.
Folkins, J. W., Moon, J. B., Luschei, E. S., Robin, D. A., Tye-Murray, N., & Moll, K. L. (1995).
What can non-speech tasks tell us about speech motor disabilities? Journal of Phonetics,
23, 139-147.
Forrest, K. (2002). Are oral-motor exercises useful in the treatment of phonological/articulatory
disorders? Seminars in Speech and Language, 23(1), 15–25.
Fozard, J. L., Vercruyssen, M., Reynolds, S. L., Hancock, P. A., & Quilter, R. E. (1994). Age
differences and changes in reaction time: The Baltimore Longitudinal Study of Aging.
Journal of Gerontology: Psychological Sciences, 49, 179–189.
Fraser, S. A., Li, K. Z. H., & Penhune, V. B. (2009). A comparison of motor skill learning and
retention in younger and older adults. Experimental Brain Research, 195(3), 419-27.
Freedman, S. E., Maas, E., Caligiuri, M. P., Wulf, G., & Robin, D. A. (2007). Internal vs.
external: Oral motor performance as a function of attentional focus. Journal of Speech,
Language, and Hearing Research, 50, 131–136.

French, K. E., Rink, J. E., & Werner, P. H. (1990). Effects of contextual interference on retention
of three volleyball skills. Perceptual and Motor Skills, 71, 179-186.

Galligan, F. et al. (2000). Acquiring Skill. In F. Galligan et al. (Ed.), Advanced PE for Edexcel.
(1st ed., pp. 102-108). Bath: Bath Press.
Gentile, A. M. (1972). A working model of skill acquisition with application to teaching. Quest,
17, 3-23.
Goberman, A., & Coelho, C. (2002). Acoustic analysis of Parkinsonian speech II: Levodopa
related fluctuations and methodological issues. Neurorehabilitation, 17, 247-254.
Goetz, C. G., Fahn, S., Martinez-Martin, P., ......... LaPelle, N. (2007). Movement Disorder
Society-sponsored revision of the unified Parkinson's disease rating scale (MDS-
UPDRS): process, format, and clinimetric testing plan. Movement Disorders, 22, 41–47.
Goetz, C. G., Poewe, W., Rascol, O., ......... Seidl, L. (2004). Movement Disorder Society Task
Force Report on the Hoehn and Yahr Staging Scale: Status and Recommendations.
Movement Disorders, 19(9), 1020–1028.

161
Goode, S. L., & Magill, R. A. (1986). The contextual interference effects in learning 3
badminton serves. Research Quarterly for Exercise and Sport, 57, 308 - 314.
Goozee, J. V., Stephenson, D. K., Murdoch, B. E., Darnell, R. E., & Lapointe, L. L. (2005).
Lingual kinematic strategies used to increase speech rate: comparison between younger
and older adults. Clinical Linguistics & Phonetics, 19, 319–334.
Grey, M. J. (2010). Proprioceptive Sensory Feedback. In: Encyclopedia of Life Sciences (ELS).
John Wiley & Sons, Ltd: Chichester.
Grimme, B., Fuchs, S., Perrier, P., & Schöner, G. (2011). Limb versus speech motor control: a
conceptual review. Motor Control, 15(1), 5-33.
Guadagnoli, M. A., & Kohl, R. M. (2001). Utilization of know;edge of results for motor
learning. Journal of Motor behaviour, 33(2), 217-224.
Guadagnoli, M. A., & Lee, T. D. (2004). Challenge point: A framework for conceptualizing the
effects of various practice conditions in motor learning. Journal of Motor Behaviour, 36,
212–224.
Guiard, Y. (1987). Asymmetric division of labour in human skilled bimanual action: The
kinematic chain as a model. Journal of Motor Behaviour, 19(4), 486-517.

Gunning-Dixon, F.M., & Raz, N. (2000). The cognitive correlates of white matter
abnormalities in normal aging: A quantitative review. Neuropsychology, 14, 224-
232.

Haith, A. J. C., Mial, R., & Vijayakumar, S. (2008).Unifying the sensory and motor components
of sensorimotor adaptation. Advances in Neural Information Processing Systems, 21,
593-600.
Hart, S. G. & Staveland, L. E. (1988). Development of NASA-TLX (Task Load Index): Results
of empirical and theoretical research. In P. A. Hancock & N. Meshkati (Eds.), Human
Mental Workload (pp. 139-184). North-Holland: Elsevier Science Publishers.
Hartman, D. E., & Danhauer, J. L. (1976). Perceptual features of speech for males in four
perceived age decades. Journal of the Acoustical Society of America, 59(3), 713-715.

Hasher, L., & Zacks, R. T. (1988). Working memory, comprehension, and aging: A review and a
new view. In G. H. Bower (Ed.), The Psychology of Learning and Motivation, Vol. 22
(pp. 193-225). New York, NY: Academic Press.

Hasher, L., Zacks, R. T., & Rahhal, T. A. (1999). Timing, instructions, and inhibitory control:
Some missing factors in the age and memory debate. Gerontology, 45, 355-357.

Henry, F. M., & Rogers, D. E. (1960). Increased response latency for complicated movements
and a “memory drum” theory of neuromotor reaction. Research Quarterly, 31, 448-458.

162
Ho, A. K., Iansek, R., Marigliani, C., Bradshaw, J. L., & Gates, S. (1998). Speech impairment in
a large sample of patients with Parkinson’s disease. Behavioural Neurology, 11, 131–
137.
Hodge, M. M. (2002). Non-speech oral motor treatment approaches for dysarthria: Perspectives
on a controversial clinical practice. Perspectives in Neurophysiology and Neurogenic
Speech Disorders, 12 (4), 22-28.
Hoit, J. D., & Hixon, T. J. (1987). Age and speech breathing. Journal of Speech and Hearing
Research, 30, 351–366.
Hula, S. N. A., Robin, D. A., Maas, E., Ballard, K. J., & Schmidt, R. A. (2008). Effects of
feedback frequency and timing on acquisition, retention, and transfer of speech skills in
acquired apraxia of speech. Journal of Speech Language and Hearing Research, 51(5),
1088–1113.
Jacewicz, W., Fox, R. A., O’Neill, C., & Salmons, J. (2009). Articulation rate across dialect, age,
and gender. Language Variation and Change, 21, 233–256.
Jelsma, O., & Pieters, J. M. (1989). Instructional strategy effects on the retention and transfer of
procedures of different difficulty level. Acta Psychologica 70, 219-234.
Jost, K., Bryck, R.B., Vogel, E.K.., & Mayr, U. (2011). Are old adults just like low working
memory young adults? Filtering efficiency and age differences in visual working
memory. Cerebral Cortex, 21(5), 1147-1154.
Kaipa, R., Robb. M. P., Beirne, G. A. O., & Allison, R. (2012). Recovery of speech following
total glossectomy: An acoustic and perceptual appraisal. International Journal of Speech-
Language Pathology.14(1), 24-34

Katz, W. F., McNeil, M. R., & Garst, D. M. (2010). Treating apraxia of speech (AOS) with
EMA-supplied visual augmented feedback. Aphasiology, 24(6–8), 826–837.
Kawashima, R., Tajima, N., Yoshida, H., Okita, K., Sasaki, T., Schormann, T., Ogawa, A.,
Fukuda, H., & Zilles, K. (2000). The effect of verbal feedback on motor learning—a PET
study. Neuroimage, 12, 698–706.
Kahneman, D. (1973). Attention and effort. Englewood Cliffs, NJ: Prentice Hall.
Keele, S. W. (1968). Movement control in skilled motor performance. Psychological Bulletin,
70, 387-403.
Kelso, J. A. S. (1992). Theoretical concepts and strategies for understanding perceptual-motor
skill: From information capacity in closed systems to self organization in open,
nonequilibrium systems. Journal of Experimental Psychology: General, 121, 26-261.
Kent, R., & Rosenbek, J. C. (1982). Prosodic disturbance and neurologic lesion. Brain and
Language, 15, 259-291.

163
Keramati, M., Dezfouli, A., & Piray, P. (2011). Speed/accuracy trade-off between the habitual
and the goal-directed processes. PLoS Computational Biology, 7(5), 1-21.
Kernodle, M. W., & Carlton, L. G. (1992). Information feedback and the learning of multiple-
degree-of-freedom activities. Journal of Motor Behaviour, 24, 187-196.
Kerr, R, & Booth, B. (1978). Specific and varied practice of a motor skill. Perceptual and Motor
Skills, 46, 395-401.
Kester, J. D., Benjamin, A. S., Castel, A. D., & Craik, F. I. M. (2002). Memory in the elderly. In
A. D. Baddeley, M. D. Kopelman, & B. A. Wilson (Eds.), Handbook of memory
disorders (2nd ed., pp. 543-567). New York: Wiley.
Ketcham, C. J., & Stelmach, G. E. (2001). Age-related declines in motor control. In J. Birren &
K. W. Schaie (Eds.), Handbook of psychology of aging (5th ed., pp. 313–348). San
Diego, CA: Academic Press.
Knapp, B. (1967). Skill in Sport: The Attainment of Proficiency. London: Routledge & Kegan
Paul.
Knapp, C. G., & Dixon, W. R. (1952). Learning to juggle: II. A study of whole and part methods.
Research Quarterly, 23, 398-401.
Knock, T. R., Ballard, K. J., Robin, D. A., & Schmidt, R. A. (2000). Influence of order of
stimulus presentation on speech motor learning: A principled approach to treatment for
apraxia of speech. Aphasiology, 14(5–6), 653–668.

Kwakkel, G. (2009). Intensity of practice after stroke : More is better. Schweizer Archiv fur
Neurologie und Psychiatrie, 160, 295-8

Laguna, P. (2008). Task complexity and sources of task-related information during the
observational learning process. Journal of Sports Sciences, 26 (10), 1097-1113.
Lai, Q., & Shea, C. H. (1998). Generalized motor program (GMP) learning: Effects of reduced
frequency of knowledge of results and practice variability. Journal of Motor Behaviour,
30, 51–59.
Lai, Q., Shea, C. H., Wulf, G., & Wright, D. L. (2000). Optimizing generalized motor program
and parameter learning. Research Quarterly for Exercise and Sport, 71, 10–24.
Lass, N. J., & Pannbacker, M. (2008). The application of evidence-based practice to non-speech
oral motor treatments. Language, Speech, and Hearing Services in Schools, 39, 408–421.
Latash, L. M., & Mikaelian, I. L. (2011). A logarithmic speed-difficulty trade-off in speech
production. Motor Control, 15, 52-67.
Latash, M. (1998). Progress in Motor Control: Bernstein's Traditions in Movement Studies
(Vol. 1). Champaign, IL: Human Kinetics.

164
Lee, T. D., & Magill, R. A. (1983). The locus of contextual interference in motor-skill
acquisition. Journal of Experimental Psychology: Learning, Memory, and Cognition, 9,
730–746.
Lee, T. D., Magill, R. A., & Weeks, D. J. (1985). Influence of practice schedule on testing
schema theory predictions in adults. Journal of Motor Behaviour, 17, 283–299.

Lin, C. H., Sullivan, K. J., Wu , A. D., Kantak, S., & Winstein C. J. (2007). Effect of task
practice order on motor skill learning in adults with Parkinson disease: a pilot study.
Physical Therapy, 87(9), 1120-1131.
Lof, G. L. (2006, November). Logic, theory and evidence against the use of non-speech oral
motor exercises to change speech sound productions. Paper presented at the annual
meeting of the American Speech, Language, and Hearing Association, Miami Beach, FL.
Lof, G. L., & Watson, M. (2008). A nationwide survey of non-speech oral motor exercise use:
Implications for evidence based practice. Language, Speech, and Hearing Services in
Schools, 39, 392–407.
Logemann, J. A., Fisher, H., Boshes, B., & Blonsky, E. R. (1978). Frequency and concurrence of
vocal tract dysfunctions in the speech of a large sample of Parkinson patients. Journal of
Speech and Hearing Disorders, 42, 47–57.
Ludlow, C., Connor, N., & Bassich, C. (1987). Speech timing in Parkinson’s and Huntington’s
disease. Brain and Language, 32, 195–214.
Maas, E., & Farinella, K. A. (2012). Random Versus Blocked Practice in Treatment for
Childhood Apraxia of Speech. Journal of Speech, Language, and Hearing Research, 55,
561-578.
Maas, E., Barlow, J., Robin, D., & Shapiro, L. (2002). Treatment of sound errors in aphasia and
apraxia of speech: Effects of phonological complexity. Aphasiology, 16(4–6), 609–622.

Maas, E., Robin, D. A., Hula, S. N. A., Freedman, S. E., Wulf, G., Ballard, K. J., & Schmidt, R.
A. (2008). Principles of motor learning in treatment of motor speech disorders. American
Journal of Speech-Language Pathology, 17, 277–298.

Mackenzie, C., Muir, M., & Allen, C. (2010). Non-speech oro-motor exercise use in acquired
dysarthria management: Regimes and rationales. International Journal of Language and
Communication Disorders, 45, 617-629.
Magill, R. A. (2004). Motor learning and control: Concepts and applications (7th ed.). New
York: McGraw-Hill.

Magill, R. A., & Hall, K. G. (1990). A review of the contextual interference effect in motor skill
acquisition. Human Movement Science, 9, 241-289.
Manes, J., & Robin, D. A. (2012). A motor learning perspective for optimizing intervention
intensity. International Journal of Speech-Language Pathology, 14, 447 – 450.
165
Marshal, F., Bund, A., & Wiemeyer, J. (2007). Does frequent augmented feedback really
degrade learning? A meta-analysis. Bewegung und Training, 1, 74-85.
Marsden, C. D. (1989). Slowness of movement in Parkinson’s disease. Movement Disorders,
4(Suppl. 1), 26–37.
Maslovat, D., Chua, R., Lee, T. D., & Franks, I. M. (2004). Contextual interference: single task
versus multitask learning. Motor Control, 8, 213-233.
Mattay, V. S., Fera, F., Tessitore, A., Hairi, A. R., Das, S., Callicott, J. H., & Weinberger, D. R.
(2002). Neurophysiological correlates of age-related changes in human motor function.
Neurology, 58, 630-635.
McCauley, R. J. , Strand, E., Lof, G. L., Schooling, T., Frymark, T. (2009). Evidence-based
systematic review: Effects of non-speech oral motor exercises on speech. American
Journal of Speech-Language Pathology , 18, 343–360.
McLeod, S., Harrison, L. J., McAllister, L. & McCormack, J. (2007, November). Prevalence of
speech and language impairment in 4,983 four-five-year-old Australian children. Poster
presented at American Speech-Language-Hearing Association Convention, Boston.
McNevin, N., Shea, C. H., Wulf, G. (2003). Increasing the distance of an external focus of
attention enhances learning. Psychological Research, 67, 22-29.
Meira Jr., C. M., & Tani, G. (2001). The contextual interference effect in acquisition of dart
throwing skill tested on a transfer test with extended trials. Perceptual and Motor Skills,
92, 910-918.
Melcher, R. T. (1934). Children’s motor learning with and without vision. Child Development, 5,
315-350.

Ménard, L., Polak, M., Denny, M., Burton, E., Lane, H., Mathies, M. L., Perkell, J. S., Tiede, M.,
& Vick, J. (2007). Interactions of speaking condition and auditory feedback on vowel
production in postlingually deaf adults with cochlear implants. Journal of Acoustical
Society of America, 121, 3790–3801

Murray, E., McCabe, P., & Ballard, K. (2011). Using rest intervention for paediatric cerebellar
ataxia [Abstract]. In B. Maassen & P. van Lieshout (Eds.). Proceedings of the 6th
International Conference on Speech Motor Control, 17, 55.
Murray, S. R., & Udermann, B. E. (2003). Massed versus Distributed Practice: Which is better?
Colorado Association for Health, Physical Education, Recreation, and Dance Journal,
28(1), 19-22.
Myerson, J., Hale, S., Hirschman, R., Hansen. C, & Christiansen, B. (1989). Global increase in
response latencies by early middle age: Complexity effects in individual performances.
Journal of the Experimental Analysis of Behaviour, 52, 353-362.

166
Myerson, J., Hale, S., Wagstaff, D., Poon, L. W, & Smith, G. A. (1990). The information-loss
model: A mathematical theory of age-related cognitive slowing. Psychological Review,
97, 475-487.
Naylor, J., & Briggs, G. (1963). Effects of task complexity and task organization on the relative
efficiency of part and whole training methods. Journal of Experimental Psychology, 65,
217-244.
Netsell, R. (1983). Speech motor control: Theoretical issues with clinical impact. In W. R. Berry
(Ed.), Clinical Dysarthria. 1-19. San Diego: College Hill.

Netsell, R. (1986). A neurobiologic view of speech production and the dysarthrias. San Diego:
College-Hill.
Nettlebeck, T., & Kirby, N. H. (1976). A comparison of part and whole training methods with
mildly mentally retarded workers. Journal of Occupational Psychology, 49, 115-120.
Newell, K. M., Carlton, M. J., & Antoniou, A. (1990). The interaction of criterion and feedback
information in learning a drawing task. Journal of Motor Behaviour, 22, 536–552.
Oldfield, R. C. (1971). The assessment and analysis of handedness: The Edinburgh Inventory.
Neuropsychologia, 9, 97–113.
Packman, A., & Onslow, M. (2012). Investigating optimal intervention intensity with the
Lidcombe Program of early stuttering intervention. International Journal of Speech-
Language Pathology, 14, 467-470.
Park, J. H., Wilde, H., & Shea, C. H. (2004). Part-whole practice of movement sequences.
Journal of Motor Behaviour, 36, 51–61.
Parnell, M. M., & Amerman, J. D. (1996). An 11-year follow-up of motor speech abilities in
elderly adults. Clinical Linguistics &Phonetics, 10, 103–118.
Partridge, L., & Mangel, M. (1999). Messages from mortality: the evolution of death rates in the
old. Trends in Ecology & Evolution, 14, 438–442.
Pendt, L. K., Reuter, I., & Müller, H. (2011). Motor Skill Learning, Retention, and Control
Deficits in Parkinson's Disease. PLoS ONE, 6(7): e21669, 1-10.
Perrier, P., Ostry, D. J., & Laboissière, R. (1996): The Equilibrium Point Hypothesis and its
application to speech motor control. Journal of Speech and Hearing Research, 39, 365–
378.
Perkell, J. S., Guenther, F. H., Lane, H., Matthies, M. L., Perrier, P., Vick, J., Wilhelms-
Tricarico, R. & Zandipour, M. (2000). A theory of speech motor control and supporting
data from speakers with normal hearing and with profound hearing loss. Journal of
Phonetics, 28, 233-372.

167
Perkell, J. S., Matthies, M. L., Lane, H., Guenther, F. H., Wilhelms-Tricarico, R., Wozniak, J.,&
Guiod, P. (1997). Speech motor control: Acoustic segmental goals, saturation effects,
auditory feedback and internal models. Speech Communication, 22, 227-250.
Perrot, A., & Bertsch, J. (2007). Role of age in relation between two kinds of abilities and
performance in acquisition of new motor skill. Perceptual Motor Skills, 104, 91–101.
Pew, R. W. (1974). Human perceptual-motor performance. In B. H. Kantowitz (Ed.), Human
information processing: Tutorials in performance and cognition. New York: Erlbaum.
Poole, J. L. (1991). Applications of motor learning principles in occupational therapy. American
Journal of Occupational Therapy, 45, 531-537.
Powell, T. (2008). An integrated evaluation of non-speech oral motor treatments. Language,
Speech, and Hearing Services in Schools, 39, 422–427.
Prado, E., & Ullman, M. T. (2009). Can imageability help us draw the line between storage and
composition? Journal of Experimental Psychology: Language, Memory, and Cognition,
35(4), 849–866.
Rohrer, D., & Taylor, K. (2006). The effects of overlearning and distributed practice on the
retention of mathematics knowledge. Applied Cognitive Psychology, 20, 1209–1224.
Rohrer, D., Taylor, K., Pashler, H., Wixted, J. T, & Cepeda, N. J. (2005). The effect of
overlearning on long-term retention. Applied Cognitive Psychology, 19, 361–374.
Rosenbek, J. C., Lemme, M. L., Ahern, M. B., Harris, E. H., & Wertz, R. T. (1973). A treatment
for apraxia of speech in adults. Journal of Speech and Hearing Disorders, 38, 462–472.
Rostami, H. R., & Ashayeri, H. (2009). Effects of motor skill practice on reaction time and
learning retention in Parkinson’s disease. Neurology India, 57(6), 768-771.
Roy, N. (2012). Optimal dose-response relationships in voice therapy. International Journal of
Speech-Language Pathology, 14, 419–423.
Ruscello, D. M. (2008). Non-speech oral motor treatment issues related to children with
developmental speech sound disorders. Language, speech, and hearing services in
schools, 39, 380–391.
Ryan, W. J. (1972). Acoustic aspects of the aging voice. Journal of Gerontology, 27, 265–268.
Sadagopan, N. (2008). Novel Nonword Learning in Older Speakers. (Unpublished doctoral
dissertation). Purdue University, West Lafayette.
Salman. M. (2002). The Cerebellum: It’s about time! But timing is not everything-New insights
into the role of the cerebellum in timing motor and cognitive tasks. Journal of Child
Neurology, 17(1), 1-9.
Salmoni, A. W., Schmidt, R. A., & Walter, C. B. (1984). Knowledge of results and motor
learning: A review and critical reappraisal. Psychological Bulletin, 95, 355–386.

168
Salthouse, T. A. (1985). A Theory of Cognitive Aging. Amsterdam: North-Holland

Saunders, J., & Knill, D. C. (2005). Humans use continuous visual feedback from the hand to
control both the direction and distance of pointing movements. Experimental Brain
Research, 162(4), 458-73.

Schmall, V. L. (1993). Sensory changes in later life. Pacific Northwest Extension publication
PNW 246, Covallis, OR: Oregon State University.
Schmidt, R. A. (1975). A schema theory of discrete motor skill learning. Psychological Review,
82, 225–260.
Schmidt, R. A. (1991). Motor learning and performance: from principles to practice.
Champaign, IL: Human Kinetics Books.
Schmidt, R. A. (2003). Motor schema theory after 27 years: Reflections and implications for a
new theory. Research Quarterly for Exercise and Sport, 74, 366–375.
Schmidt, R. A., & Lee, T. D. (2005). Motor control and learning: A behavioral emphasis (4th
ed.). Champaign, IL: Human Kinetics.
Schmidt, R. A., & White, J. L. (1972). Evidence for an error detection mechanism in motor
skills: A test of Adams' closed-loop theory. Journal of Motor Behaviour, 4, 143-153.
Schmidt, R. A., & Wrisberg, C. A. (1973). Further tests of the Adams' closed-loop theory:
Response produced feedback and the error-detection mechanism. Journal of Motor
Behavior, 3, 155-164.
Schmidt, R. A., Young, D. E., Swinnen, S., & Shapiro, D. C. (1989). Summary knowledge of
results for skill acquisition: Support for the guidance hypothesis. Journal of Experimental
Psychology: Learning, Memory, and Cognition, 15(2), 352-359.
Schulz, G. M., Stein, L., & Micallef, R. (2001). Speech motor learning: preliminary data.
Clinical linguistics & Phonetics, 15, 157-161.
Searl, J. P., Gabel, R. M., & Fulks, J. S. (2002). Speech disfluency in centenarians. Journal of
Communication Disorders, 35, 383–392.
Seidler, R. A. (2006). Differential effects of age on sequence learning and sensorimotor
adaptation. Brain Research Bulletin, 70, 337–346.
Sekiya, H., Magill, R. A., & Anderson, D. I. (1996). The contextual interference effect in
parameter modifications of the same generalized motor program. Research Quarterly for
Exercise and Sport, 67, 59-68.
Shea, C. H., & Kohl, R. M. (1991). Composition of practice: Influence on the retention of motor
skills. Research Quarterly for Exercise and Sport, 62, 187–195.

169
Shea, C. H., Lai, Q., Wright, D. L., Immink, M., & Black, C. (2001). Consistent and variable
practice conditions: Effects on relative and absolute timing. Journal of Motor Behaviour,
33, 139–152.
Shea, C. H., Wulf, G., Park, J. H., & Gaunt, B. (2001). Effects of an auditory model on the
learning of relative and absolute timing. Journal of Motor Behaviour, 33, 127-138.
Shea, J. B., & Morgan, R. L. (1979). Contextual interference effects on the acquisition, retention,
and transfer of a motor skill. Journal of Experimental Psychology: Human Learning &
Memory, 5, 179–187.
Shea, J. B., & Wright, D. L. (1991). When forgetting benefits motor retention. Research
Quarterly for Exercise and Sport, 62, 293-301.
Sherwood, D. E. (1996). The benefits of random variable practice for spatial accuracy and error
detection in a rapid aiming task. Research Quarterly for Exercise and Sport, 67, 35-43.
Sherwood, D. E., & Lee, T. D. (2003). Schema theory: Critical review and implications for the
role of cognition in a new theory of motor learning. Research Quarterly for Exercise and
Sport, 74, 376–382.
Shimokata, H., & Kuzuya, F. (1995). Two-point discrimination test of the skin as an index of
sensory aging. Gerontology, 41, 267-272.

Shinomori, K., Schefrin, B. E., & Werner, J. S. (2001). Age-related changes in wavelength
discrimination. Journal of the Optical Society of America A, 18, 310–318.

Shoenfelt, E. L., Synder, L. A., Maue, A. E., McDowell, C. P., & Woolard, C. D. (2002)
Comparison of constant and variable practice conditions on free-throw shooting.
Perceptual Motor Skills, 94, 1113- 1123.
Shumway-Cook, A., & Woollacott. M. (2000). Motor Control: Theory and Practical
Applications (2nd ed). Baltimore, MD: Lippincott, Williams and Wilkens.

Sidaway, B., Gordon, R., Hopkins, M., Kershaw, M., Marean, C., & Wilkins, N. (2006). Random
and blocked practice schedule effects on motor skill learning in individuals with
Parkinson’s disease [abstract]. Journal of Neurologic Physical Therapy, 30(4), 204-205.
Skinder-Meredith, A. E., & Lentz, L. A. (2004). Effect of an oral motor exerciser in articulation
therapy: A single subject design. Retrieved July 4, 2012, from
http://www.speechpathology.com/articles/arc_disp.asp?id=237.
Smith, A. (2006). Speech motor development: Integrating muscles, movements, and linguistic
units. Journal of Communication Disorders, 39, 331–349.
Soliveri, P., Brown, R. G., Jahanshahi, M., & Marsden, C. D. (1992). Effect of practice on
performance of a skilled motor task in patients with Parkinson’s disease. Journal of
Neurology, Neurosurgery & Psychiatry, 55(6), 454–460.

170
Steinhauer, K., & Grayhack, J. P. (2000). The role of knowledge of results in performance and
learning of a voice motor task. Journal of Voice, 14(2), 137–145.
Strickgold, R., & Walker, M. P. (2005). Memory consolidation and reconsolidation: what is the
role of sleep? Trends in Neurosciences, 28, 408–415.
Svinin, M., Riken, N., Goncharenko, I., Hosoe, S., & Kanou, Y. (2007). Modeling and motor
learning of reaching movements in virtual environments. Proceedings of the
International Conference on Complex Medical Engineering, IEEE/ICME held in
Shanghai, May 23-27th, 2007. (pp. 1277-1284). Hong Kong: K. C. Wong Education
Foundation.

Swinnen, S., Schmidt, R. A., Nicholson, D. E., & Shapiro, D. C. (1990). Information feedback
for skill acquisition: Instantaneous knowledge of results degrades learning. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 16(4), 706–716.

Swinnen, S.,Walter, C. B., Lee, T. D., & Serrien, D. J. (1993). Acquiring bimanual skills:
Contrasting forms of information feedback for interlimb decoupling. Journal of
Experimental Psychology: Learning, Memory, and Cognition, 19, 1328–1344.
Takeda, N., Thomas, G. R., & Ludlow, C. I. (2000). Aging effects on motor units in the human
thyroarytenoid muscle. Laryngoscope, 110, 1018–1025.
To, C. K. S., Law, T., & Cheung, P. S. P. (2012). Treatment intensity in everyday clinical
management of speech sound disorders in Hong Kong. International Journal of Speech-
Language Pathology, 14, 462–466.
Tremblay, K., Piskosz, M., & Souza, P. (2003). Effects of age and age-related hearing loss on the
neural representation of speech cues. Clinical Neurophysiology, 114, 1332–1343.
Vance, J., Wulf, G., Tollner. T., McNevin, N., & Mercer. J. (2004). EMG activity as a function
of the performer's focus of attention. Journal of Motor Behaviour, 36(4), 450-459.
Verhoff, J., Reuman, D., & Feld, S. (1984). Motives in American men and women across the
adult life span. Developmental Psychology, 20, 1142-1158.
Verschueren, S. M. P., Swinen, S. P., Dom, R., & DeWeerdt, W. (1997). Interlimb coordination
in patients with Parkinson’s disease: Motor learning deficits and the importance of
augmented information feedback. Experimental Brain Research, 113(3), 497-508.
Vickers, J. N. (1994). The reversal effect. In A. Jewett, L. Bain & C. Ennis (Ed). The Curriculum
Process in Physical Education. 348-353. Madison, Wisconsin: Brown & Benchmark.
Voelcker-Rehage, C. (2008). Motor-skill learning in older adults—A review of studies on age-
related differences. European Review of Aging and Physical Activity, 5, 5–16.
Walker, M. P., Brakefield, T., Morgan, A., Hobson, J. A., & Strickgold, R. (2002). Practice with
sleep makes perfect: sleep-dependent motor skill learning. Neuron, 35, 205–211.

171
Webb, W., & Adler, R. K. (2008). Neurology for the Speech-Language Pathologist (5th ed.).
Mosby Elsevier Health Sciences: PA, USA.
Weeks, D. L., & Kordus, R. N. (1998). Relative frequency of knowledge of performance and
motor skill learning. Research Quarterly for Exercise & Sport, 69, 224-230.
Weismer, G. (2006). Philosophy of research in motor speech disorders. Clinical Linguistics and
Phonetics. 20(5), 315-349.
Welford, A.T. (1985). Changes of Performance with Age: an overview. In: N. Charne (Ed.).
Aging and Human Performance (pp. 333 – 369), Chichester: Wiley.
Wickelgren, W.A. (1977). Speed–accuracy tradeoff and information processing dynamics. Acta
Psychologica, 41, 67–85.
Wickstorm, R. L. (1958). Comparative study of methodologies for teaching gymnastics and
tumbling stunts. Research Quarterly, 29, 109-115.
Wilson, E., Green, J., Yunusova, Y., & Moore, C. (2008). Task specificity in early oral motor
development. Seminars in Speech and Language, 29, 257–266.
Wilson, S. A. K. (1925). Disorders of motility and muscle tone, with special reference to the
striatum. Lancet, 2, 1–53, 169, 215, 268.
Wing, A. M. (2002). Voluntary timing and brain function: An information processing approach.
Brain and cognition, 48, 7-30.
Wing, A. M., & Kristofferson, A. B. (1973b). Response delays and the timing of discrete motor
responses. Perception & Psychophysics, 14, 5-12.
Winstein, C. J. (1991). Knowledge of results and motor learning. Implications for physical
therapy. Physical Therapy, 71,140–149.
Winstein, C. J., & Schmidt, R. A. (1990). Reduced frequency of knowledge of results enhances
motor skill learning. Journal of Experimental Psychology: Learning, Memory, and
Cognition, 16, 677–691.

Wishart, L. R., Lee, T.D., Cunningham, S. J., & Murdoch, J.E. (2002). Age- related differences
in the role of augmented visual feedback in learning a bimanual coordination pattern.
Acta Psychologica, 110, 247-263.
Wong, A. Y. H., Ma, E. P. M., & Yiu, E. M. L. (2011). Effects of Practice Variability on
Learning of Relaxed Phonation in Vocally Hyperfunctional Speakers. Journal of Voice,
25(3), e103-113.
Wright, B. M., & Payne, R. B. (1985). Effects of aging on sex differences in psychomotor
reminiscence and tracking proficiency. Journal of Gerontology, 40, 179-184.

172
Wright, D. L., Black, C. B., Immink, M. A., Brueckner, S., & Magnuson, C. (2004). Long-term
motor programming improvements occur via concatenating movement sequences during
random but not blocked practice. Journal of Motor Behaviour, 36, 39–50.
Wright, D. L. (1991). The role of intertask and intratask processing in acquisition and retention
of motor skills. Journal of Motor Behaviour, 23, 139-145.
Wright, D. L. (1991). The role of intertask and intratask processing in acquisition and retention
of motor skills. Journal of Motor Behaviour, 23, 139-145.
Wulf, G. (2007). Attention and motor skill learning. Champaign, IL: Human Kinetics.
Wulf, G, Chiviacowsky, S., Schiller, E., & Ávila, L. T. G. (2010). Frequent external-focus
feedback enhances motor learning. Frontiers in Movement Science and Psychology, 1,
190.
Wulf, G., & Lee, T. D. (1993). Contextual interference in movements of the same class:
Differential effects on program and parameter learning. Journal of Motor Behaviour, 25,
254–263.
Wulf, G., & McNevin, N. (2003). Simply distracting learners is not enough: More evidence for
the learning benefits of an external focus of attention. European Journal of Sport Science,
3(5), 1–13.
Wulf, G., & Schmidt, R. A. (1997). Variability of practice and implicit motor learning. Journal
of Experimental Psychology: Learning, Memory, and Cognition, 23, 987–1006.

Wulf, G., & Shea, C.H. (2002). Principles derived from the study of simple motor skills do not
generalize to complex skill learning. Psychonomic Bulletin and Review, 9, 185-211.
Wulf, G, Höß, M., & Prinz, W. (1998). Instructions for motor learning: Differential effects of
internal versus external focus of attention. Journal of Motor Behavior, 30, 169–179.
Wulf, G, Lee, T. D., & Schmidt, R. A. (1994). Reducing knowledge of results about relative
versus absolute timing: Differential effects on learning. Journal of Motor Behavior, 26,
362–369.
Wulf, G, McNevin, N., & Shea, C. H. (2001). The automaticity of complex motor skill learning
as a function of attentional focus. The Quarterly Journal of Experimental Psychology,
54A, 1143-1154.

Wulf, G., Shea, C. H., & Matschiner, S. (1998). Frequent feedback enhances complex motor skill
learning. Journal of Motor Behavior, 30, 180-192.

Yoder, P., Fey, M. E., & Warren, S. F. (2012). Studying the impact of intensity is important but
complicated. International Journal of Speech-Language Pathology, 14, 410 – 413.

173
Zacks, R. T., Hasher, L., & Li, K. Z. H. (2000). Human memory. In T. A. Salthouse & F. I. M.
Craik (Eds.), Handbook of Aging and Cognition, (2nd ed., pp. 293-357). Mahwah, NJ:
Lawrence Erlbaum.

Ziegler, W. (2003). Speech motor control is task-specific. Evidence from dysarthria and apraxia
of speech. Aphasiology, 17, 3-36.
Zubiaur, M., Ona, A., & Delgado, J. (1999). Learning volleyball serves: A preliminary study of
the effects of knowledge of performance and results. Perceptual and Motor Skills, 89,
223-232.

174
APPENDICES

175
 Appendix 1

Project information sheet

176
 PROJECT INFORMATION SHEET FOR PARTICIPANTS

Department of Communication Disorders

Research Title:

Effect of principles of motor learning on speech and non-speech-motor learning

Principal Investigator:

Ramesh Kaipa
Ph.D. Candidate, Department of Communication Disorders
University of Canterbury
Ph: 3667001 extn 4813
Mob: 0211114032
E-mail: [email protected]

Supervisors:

Michael Robb, PhD

Professor, Department of Communication Disorders
University of Canterbury
Ph: (03) 364 2987 extn 7077
Email: [email protected]

Maggie-Lee Huckabee, PhD

Senior lecturer, Department of Communication Disorders.
University of Canterbury
Van der Veer Institute for Parkinson’s and Brain Research.
66 Stewart Street, Christchurch, New Zealand
Phone: (03) 378 6070
E-mail: [email protected]

Richard Jones, PhD

Associate Professor, Department of Communication Disorders
Senior Biomedical Engineer, Department of Medical Physics and Bioengineering, CDHB
Director, Christchurch Neurotechnology Research Programme
Van der Veer Institute for Parkinson’s and Brain Research.
66 Stewart Street, Christchurch, New Zealand
Phone: (03)378 6077
E-mail: [email protected]

177
You are invited to participate in the research project entitled “Effect of principles of motor
learning on speech and non-speech-motor learning”

The aim of this project is to evaluate the effects of motor learning tasks, such as amount and type
of practice on the learning of novel speech and non-speech tasks.

Your participation in this project will involve attending three sessions on three consecutive days
at the Communication Disorders Research Facility. The first two sessions will last approximately
one hour and ten minutes each, and the third session will last approximately 15 minutes. The first
two sessions will involve learning a novel speech and a non-speech task. The speaking task will
require you to rehearse a phrase up to 100 times. The non-speech task will involve you to play a
musical tune on an electronic key board. You will rehearse this up to 100 times. On the third day,
you will return to the institute and will be required to demonstrate the speech and musical tasks
you were taught on the previous day.

The entire procedure is completely non-invasive and does not pose any hazard to your safety.
The entire study will take place at Communication Disorders Research Facility located at 19
Creyke Road, Ilam. As a token of appreciation for your participation you will be given $25
of super market vouchers at the end of your participation. In addition, your valuable
participation will be very useful to investigate new management protocols to help treat speech
deficits in individuals with Parkinson’s disease.

You have the right to withdraw from the project at any time, including withdrawal of any
information provided. The results of the project may be published, but you may be assured of the
complete confidentiality of data gathered in this investigation: the identity of participants will not
be made public without their consent. To ensure confidentiality, the information gathered will be
assigned a number and all identifiable information removed. Data will be kept in a locked filing
cabinet within a lockable room in the Department of Communication Disorders. A summary of
the results of the study will be provided upon request.

The project is being carried out as a requirement for a PhD (Doctor of Philosophy) thesis by
Ramesh Kaipa under the supervision of Professor Michael Robb, Dr Maggie Lee Huckabee and
Assoc Prof Richard Jones. The project has been reviewed and approved by the University of

178
Canterbury Human Ethics Committee. If you have any further questions about the research
project, please do not hesitate to contact either my supervisor or myself at the University of
Canterbury. Thank you once again.

If you have any questions or concerns about your rights as a participant in this research study,
you can contact an independent health and disability advocate. This is a free service provided
under the Health and Disability Commissioner Act.

Telephone (NZ wide): 0800 555 050. Free Fax (NZ wide): 0800 2787 7678 (0800 2 SUPPORT)

Email (NZ wide): [email protected]

Sincerely,

Ramesh Kaipa B.Sc, MASLP Professor Michael Robb

PhD Student Department of Communication Disorders
Ph: 3667001 extn 4813 Ph: 364 2987 extn 7077
Mob: 0211114032 Email: [email protected]
Email: [email protected]

Department of Communication Disorders

University of Canterbury Private Bag 4800, Christchurch 8020, New Zealand.
Tel: +64 3 364-2987 x7077, Fax: +64 3 364 2260
www.cmds.canterbury.ac.nz

179
 Appendix 2

Participant consent form

180
Ramesh Kaipa
Department of Communication Disorders
University of Canterbury
Private Bag 4800
Christchurch

Consent Form

Effect of principles of motor learning on speech and non-speech-motor learning

I have read and understood the description of the above-named project. On this basis, I agree
to take part as a participant in the project, and I consent to publication of the results of the
project with the understanding that anonymity will be preserved.

I understand also that I may at any time withdraw from the project, including withdrawal of any
information I have provided.

NAME (please print): …………………………………………………………….

Signature:

Date:

Department of Communication Disorders

University of Canterbury Private Bag 4800, Christchurch 8020, New Zealand.
Tel: +64 3 364-2987 x7077, Fax: +64 3 364 2260
www.cmds.canterbury.ac.nz

181
 Appendix 3

Edinburgh Handedness Inventory (Oldfield, 1971). The 10 activities used for

the assessing the hand dominance and the scoring procedure are shown.

182
 The Edinburgh Handedness Inventory requires answers to questions about the participants’
practice in performing a number of habitual everyday activities in which the roles of right and
left hands are clearly distinguished. The inventory consists of 10 different everyday activities and
also a right and a left column for the participants to indicate their preference of handedness for
the activities. The participants were provided with the following instructions to help them
complete the inventory: “Please indicate your preference with regards to use of your hands in the
following activities by putting a check in the appropriate column. Where the preference is so
strong that you would never try to use the other hand, unless absolutely forced to, put two
checks. If in any case you are really indifferent, put a check in both columns”. The inventory is
shown in Figure 1. Scoring involved the following steps:
Calculating the total number of checks in the left and right columns and calculating the
cumulative total of the right and left total.
Calculating the difference between the right total and left total (Right total-Left total).
Dividing the “difference” by the “cumulative total” cell and multiplying by 100.

Scores below -40 are indicative of left hand dominance, scores between -40 and +40 are
indicative of ambidextrous, and scores above +40 are indicative of right hand dominance.

183
 Task / Object Left Hand Right Hand

1. Writing

2. Drawing

3. Throwing

4. Scissors

5. Toothbrush

6. Knife (without fork)

7. Spoon

8. Broom (upper hand)

9. Striking a Match (match)

10. Opening a Box (lid)

Total checks: LH = RH =

Cumulative Total CT = LH + RH =

Difference D = RH – LH =

Result R = (D / CT) 100 =

184