See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/327072132

Evaluating the impact of a generalist teacher-led music program on early

childhood school children’s singing skills and attitudes to music

Article in Psychology of Music · August 2018

DOI: 10.1177/0305735618790355

CITATIONS READS

3 407

4 authors:

Margaret Barrett Katie Zhukov

The University of Queensland The University of Queensland
106 PUBLICATIONS 1,447 CITATIONS 32 PUBLICATIONS 140 CITATIONS

SEE PROFILE SEE PROFILE

Joanne E. Brown Graham Frederick Welch

The University of Queensland University College London Institute of Education
3 PUBLICATIONS 6 CITATIONS 268 PUBLICATIONS 3,261 CITATIONS

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Sounds of Intent View project

Communities of Music Education View project

All content following this page was uploaded by Graham Frederick Welch on 22 August 2018.

The user has requested enhancement of the downloaded file.

Evaluating the impact of a generalist teacher-led music program on early
childhood school children’s singing skills and attitudes to music1

Margaret S. Barrett1, Katie Zhukov1, Joanne E. Brown2 and Graham F. Welch3

1Schoolof Music, University of Queensland, Brisbane, Queensland, Australia 2School of Psychology, University of Queensland,
Brisbane, Queensland, Australia 3Institute of Education, University College London, London, UK

Abstract

This article reports on the impact of a generalist teacher-led music program on early
childhood school children’s singing skills and attitudes to music. Singing tests and class
surveys were administered to students in 11 Australian primary schools where music
specialists mentored classroom teachers over the period of one to two school terms.
The results show that implementing music activities in early education settings can
positively impact young children’s singing skills and attitudes to music regardless of
gender, ethnicity and socio-economic standing of the school. The study provides
empirical evidence of the benefits accrued by children through access to music
education.

Key Words

early childhood, generalist music education, music development, singing, teacher

development

Introduction

The literature documenting the beneficial effects of music participation on intellectual

development of children is growing. Nevertheless, what counts as effective music
education to promote such benefits is still somewhat contested. Part of this debate
relates to the extent to which non-music specialist/generalist primary school teachers
are able to support their students’ musical development, particularly if the teachers

1
Published 16 August 2018 Psychology of Music DOI: 10.1177/0305735618790355

1
either lack sufficient music education knowledge and/or confidence to teach music
(e.g., Hennessy, 2012). This pedagogical challenge should be seen within a context
where considerable evidence from across the world is indicating that music instruction
is associated with improved measures of intellectual functioning, spatial and verbal
skills (Costa-Giomi, 2012), as well as literacy, numeracy, and aspects of executive
function (Williams et al., 2015; Sachs, Kaplan, Der Sarkissian, & Habibi, 2017). For
example, a study of 132 Canadian 6-year-olds documented greater improvement in
general IQ in students undertaking one-year tuition in keyboard and voice than those
in control groups (Schellenberg, 2004). A large survey of American children (n = 4,376)
showed that participation in music activities was associated with academic
achievement (Southgate & Roscigno, 2009). In Israel, a 2-year music enrichment
program consisting of music appreciation, performance, and creativity with 45 high-
risk elementary school children showed that children’s growing understanding of
musical structures enhanced their cognitive development in relation to general
organizing structures in other disciplines (Portowitz, Lichtenstein, Egorova, & Brand,
2009). Music listening has also been found to play a crucial role in mood maintenance
and have beneficial impact on well-being and physical health (Macdonald, 2013).
In addition to enhanced academic achievement (Brown, Benedett, &
Armistead, 2010; Hallam, 2011; Johnson & Memmott, 2006; Moreno et al., 2011;
Moritz, Yampolsky, Papadelis, Thomson, & Wolf, 2013; Rauscher & Hinton, 2011;
Ritchie & Williamon, 2011; Williams, Barrett, Welch, Abad, & Broughton, 2015) and
positive social impact (Costa-Giomi, 2004; Hallam, 2011; Rickard et al., 2013; Ritchie &
Williamon, 2011; Welch, Himonides, Saunders, Papageorgi, & Sarazin, 2014; Williams
et al., 2015), music learning is reported to produce positive music outcomes for
children, such as in the development of their singing ability and enhanced attitudes
towards music. Most recently, a major longitudinal study provides evidence of the
basis for music’s impact by examining the effects of music training on the
developmental trajectory of children’s brain structure, beginning at age six (Habibi et
al, 2017). Their evidence suggests that two years of music training induces macro- and
microstructural brain changes in school-age children.
Various interventions have been trialled to improve music skills in primary
school-age children. For example, in the USA, early childhood specialists implemented
a music curriculum after intensive professional development in music teaching,
resulting in significant improvement in children’s tonal patterns (Runfola & Etopio,
2012). Additional benefits of the music intervention in this study included increased
vocabulary and grammar understanding, particularly for students with weaker literacy
skills. Daily singing instruction for kindergarten-age children by a music specialist
enhanced accurate singing on pitch-matching tasks significantly more in comparison to

2
a control group (Demorest, Nichols & Pfordresher, 2017). However, in this study the
singing of a well-known song did not improve for either group. These results suggest
the need to consider multiple developmental and musical factors when designing age-
appropriate singing tests for children. In a slightly older age group (10–11 year olds)
singing intonation was improved through the use of gesture and movement in choral
rehearsal (Liao & Davidson, 2016). This study supports the notion of developing age-
specific music training interventions to enhance positive music outcomes for children,
an idea that has been reinforced by data from a new interview study of teachers of
singing for young children (Pecina, 2017). This emphasizes the need for individual
differentiation, as well as the use of student-led and technological feedback in the
development of young children’s singing behaviours.
The research literature also highlights gender and age differences in early
music learning, with girls and younger children demonstrating more positive attitudes
towards music in general and singing in particular (Roberts, 2016; Saunders,
Varvarigou & Welch, 2010). For example, a study of 60 English-Cantonese bilingual
children identified gender differences in singing competency, with girls outperforming
boys, particularly on singing tasks that required text recollection (Mang, 2006). This
finding is supported by a 5-year longitudinal study of American children investigating
gender differences in early childhood, in particular during transition from pre-school to
elementary school, that showed girls outperforming boys on behavioural self-
regulation (Matthews, Ponitz & Morrison, 2009). However, the study found no
significant gender differences on five academic outcomes. The available literature
seems to suggest that some gender differences may exist in early music education with
girls scoring higher on singing competency, self-efficacy, and behavioural regulation.
Developmental changes tend to occur earlier for girls than boys, with both genders
developing greater vocal pitch range by the age of 10 (Sergeant & Welch, 2009). For
young boys, observing other boys singing seems to motivate their participation and
skill development (Hall, 2005).
There is growing evidence that music learning can have significant positive
outcomes for disadvantaged children living in low socio-economic areas (Osborne et
al., 2016). Interestingly, McPherson, Osborne, Barrett, Davidson, and Faulkner (2015)
reported that children from low socio-economic backgrounds have a greater desire to
learn a musical instrument than children from better-off families. This study illustrates
that socio-economic disadvantage need not be a barrier to music learning, and echoes
findings concerning measures of actual singing behaviours in a national dataset of n =
8,162 children in England (Welch et al., 2009). Despite being in areas of multiple
deprivation, child participants within the National Singing Programme Sing Up
achieved significantly more advanced singing behaviours than comparison child

3
controls from outside the programme who were attending schools in less deprived
areas of England.
To evaluate the effectiveness of any music learning intervention on
improvement in singing skills, it is necessary to test participating children before and
after training and also to compare results to controls. Literature has shown that
accurate testing of children’s singing is possible, with researchers demonstrating high
reliability of results (Mecke & Sundberg, 2010; Nichols & Wang, 2016). However, when
designing singing tests, it is important to consider the types of items being assessed, as
item and task difficulty level are amongst various factors affecting singing voice
development (Nichols, 2016a). For example, one of the factors that positively impacts
children’s singing performance during testing is singing alone or with others (Nichols,
2016b).
Despite the steady accumulation of research evidencing the multiple benefits
of quality music education, and singing in particular, children’s access to music
education and engagement in formal schooling is variable in both extent and quality.
Within Australia despite repeated calls for increased provision of music education the
decade following the National Review of School Music Education (Department of
Education Science & Training (DEST), 2005) witnessed a continuing decline in provision
of music in school education and teacher education for generalist teachers (Letts,
2015). To remedy this situation and to enable generalist music teachers to deliver
quality music teaching in their classrooms, an intervention funded by the Federal
Government was implemented in Australia from 2015. The National Music Teacher
Mentoring Program (NMTNP), an initiative of educator Richard Gill, a Member of the
Order for Australia (AM), and funded by the Federal Ministries of Arts and Education
with contributions from relevant Ministries in NSW and Victoria, has provided a
mechanism for specialist music educators to mentor generalist early childhood
(kindergarten to year 2, ages 4 –8) classroom educators over an extended period of
time. This article reports on the impact of this mentored generalist teacher-led music
program on early childhood school children’ singing skills and attitudes to music and
addresses the following research questions:
1. What improvements in young children’s singing are evidenced following the
implementation of the mentoring program?
2. (2) What impact did the mentoring program have on children’s attitudes
towards music?

Method

4
Design of the study

The authors evaluated the implementation of the National Music Teacher Mentoring
Program (NMTMP) in 11 Australian Primary schools in New South Wales (20-week
implementation in eight schools) and Victoria (10-week implementation in three
schools). The success of the programme was measured by pre-to-post changes in
singing of participating children and class surveys of children’s attitudes to music.
Within the participant schools there were mentored classes and non-mentored control
classes of matching year level and demographic profile to that of the experimental
groups.

Mentoring procedures

The mentoring programme began with a two-day master class that provided mentors
with a good understanding of the fundamentals of mentoring and music, and enabled
them to share simple strategies, practices and resources for engaging students in
quality music education. The experienced music educators were paired off with
generalist classroom teachers who volunteered for the program. Each mentor worked
collaboratively with their mentees to develop lesson plans and music activities that
utilized existing resources and were integrated into the daily program of the class,
often occurring during transitions between lessons.

Participants

Seven mentors worked with 19 classroom teachers. One school in the evaluation had a
resident music specialist, but with no music provision in the remaining schools. Two
hundred and ninety-two children participated in singing tests and surveys. The children
were aged 4–8 years and were drawn from early years (reception classes) and primary
school years 1 and 2. The number of children in the mentored classes was 237, with 55
children in the control classes of matched year level.

Data collection procedures

Mentors administered the singing test individually to each child in a space that was
familiar outside the classroom, with assistance from the classroom teachers, to ensure
that children were comfortable. Mentors’ musical expertise was required to assess the
quality of singing. No starting pitch was given for the two test songs. Mentors

5
completed hard copies of individual singing test ratings and these were then provided
to the researchers for data analyses. Classroom teachers administered a class survey
(see below) as a group activity with their students. If children needed support with
reading, or comprehension for individual items, this was provided by the teachers and
support staff as the survey was being completed. Researchers collated and analysed
the survey data.

Data gathering instruments

Demographics. Both singing tests and surveys included demographic data, such as date
of birth, gender, school year level, and ethnicity of children. Teachers provided this
information from the relevant school records. The age of each participant was
calculated in full years from the date of birth at the time of pre-test administration.
The Australian Bureau of Statistics (ABS) classification of countries (ABS, 2014) was
used to code the ethnicity of the participants. The ABS classifications of socio-
economic indexes by postcode (ABS, 2013) were used to code each school by its
location. A socio-economic category was assigned, based upon the region in which the
child’s school was located, rather than on an individual child basis sourced from family
income. The school’s socio-economic index was considered to be indicative of the
educational resources available at the school. The socio-economic status is divided into
10 deciles, with 1 being the most disadvantaged and 10 being the most advantaged.

Singing test piece. Two well-known simple songs were chosen as test pieces: “Twinkle,
Twinkle, Little Star” and “Happy Birthday.” The first song requires an opening pitch of a
fifth followed largely by descending stepwise movement and has a pitch range of a
sixth; the second song contains an octave leap that is more challenging to pitch
correctly and has a more complex melodic movement. Singing of each song was rated
by music mentors using a template (see Appendix A), with a score of 1 representing a
chant-like rendition, the words of the song appearing to be the initial centre of interest
rather than the melody; a score of 2 demonstrating outlines of the target melody and
singing pitch range expanding; a score of 3 given when melodic shape and intervals
were mostly accurate; and a score of 4 awarded when no significant pitch or rhythm
errors were present. In addition to the evaluation of singing as outlined above
participants’ speaking voice range was also identified. The singing test has been
validated in the evaluation of the National Singing Programme in the UK with over
11,000 children (Welch et al., 2014).

6
Student survey. The implemented survey was a shortened version of the survey
devised by Welch et al. (2014), with the original 60 questions abridged to 24, due to
the younger target population (see Appendix B). The student survey consisted of
questions regarding general attitudes towards music and singing, in addition to
demographics. The questions focused on musical context at school and home, identity
as a musician, self and social inclusion, emotional engagement, health and happiness,
and resilience. For each question, children were asked to choose one face from a range
of seven faces, ranging from sad to happy; this was later collapsed into three faces
(sad, neutral, happy) to facilitate implementation of the survey. The use of emojis was
deemed more appropriate for the target age-group of children rather than numbering
or word labels, as are typically used in survey scales.

Statistical analyses
An Excel file with results from the individual singing tests was created with school,
socio-economic index by postcode, school year level, age, gender, and ethnicity coded
for each participant. Quantitative data underwent statistical analyses using SPSS 22.0
software.
The procedures for singing test analyses included preliminary checks on
categorical variables, demographic breakdown of the participants, preliminary
analyses on continuous variables, descriptive statistics, missing data analysis, and focal
data analyses of mentored children results versus control children results from pre- to
post-test, and improvement in song 1 and song 2 from pre- to post-test. The results
were also analysed for potential socio-economic, school year level, age, gender and
ethnicity effects.
The class survey data underwent similar data cleaning and preparation
procedures as out- lined above. The overall scale demonstrated a good level of
reliability from pre- to post-assessment and was utilized as a measure of overall
general positive attitudes toward music. These scores were calculated as the mean
value of the attitude ratings across the 24 items. The survey results were analysed for
changes in general attitudes towards music from pre- to-post-mentoring, and potential
socio-economic, year level, age, gender, and ethnicity differences.

Results

Demographics

7
Table 1 reports sample demographics of children who participated in the singing tests
and surveys. As seen from the table, a disproportionately larger number of participants
came from the mentoring program compared to the control. This was due to practical
constraints concerning the recruitment of control participants of matched year level
and demographic profile to that of the intervention groups, as well as an emphasis on
collecting maximal treatment data within the limited school resources made available.
The gender breakdown of the sample provided a relatively equal split of male and
female participants. As expected in the Australian context, the number of children who
claimed Oceania ethnicity far outweighed that of the other categories. Whilst many
children had an other than English ethnic/language background, the majority were
born in Australia. Year 1 children were over-represented in the sample, while year 2
children were under-represented relative to kindergarteners. This was due to many
classes in the participating schools being composite year 1/2 classes, resulting in
greater concentration in the middle, year 1 level.

Singing tests results

Mentored versus control groups. Paired samples t-tests showed that, while children in
the control groups showed no overall difference in rated singing ability between
baseline (M = 2.01, SE = 0.17) and post-intervention (M = 2.03, SE = 0.16)
performances, t(34) = −0.17, p = .865, children in the mentored treatment groups
demonstrated significant improvement following intervention (M = 2.94, SE = 0.05)
compared to baseline, pre-intervention singing ability scores (M = 2.15, SE = 0.06),
t(190) = 16.59, p < .001 (see Figure 1). Insufficient data on children’s speaking voice
range was gathered across all groups. Consequently, these data have not been
included in this analysis.

Gender. Statistical analyses (ANCOVAs) showed no gender differences between boys

and girls on both songs.

Age and year level. Multiple regression analyses were conducted to explore the
potential developmental influence of age on singing performance. Age was found not
to be a factor in improvement for song 1. However, increased age was associated with
higher performance scores on the more difficult song 2. This result needs to be
interpreted with some caution due to very low numbers of participants in the
youngest (4 years old) and oldest (8 years old) age bracket (see Table 2).
Consequently, these findings suggest it more prudent to employ comparisons
of children’s year level to identify key developmental periods in which singing

8
education might be more effective. Overall improvement in rated singing ability for
song 1 was significantly more pronounced for year 1 children (M = 1.02, SE = 0.08) than
either their kindergarten (M = 0.49, SE = 0.08), t(222) = 4.72, p < .001, or year 2
counterparts (M = 0.49, SE = 0.11), t(222) = 3.94, p < .001 (see Figure 2). Similarly for
song 2, year 1 participants (M = 0.89, SE = 0.06) outperformed both the kindergarten
(M = 0.36, SE = 0.07), t(222) = 5.56, p < .001, and year 2 children (M = 0.47, SE = 0.10),
t(222) = 3.70, p < .001, in terms of rate of singing improvement shown from pre- to
post-intervention (see Figure 2).

Table 1. Singing tests and survey demographics by treatment group, gender, ethnicity,
year level, school and socio-economic index

Variable Categories/ Levels Frequency Percent

Treatment Group Control 55 18.84%
Treatment 237 81.16%
Gender Male 144 49.32%
Female 148 50.68%
Ethnicity Oceania 126 43.15%
North Africa or Middle East 48 16.44%
North-East Asia 44 15.07%
Southern or Central Asia 33 11.30%
South-East Asia 17 5.82%
South-Eastern Europe 10 3.42%
North-West Europe 6 2.05%
Sub-Saharan Africa 6 2.05%
Americas 2 0.68%
Year Kindergarten 92 31.51%
Year 1 125 42.81%
Year 2 75 25.68%
School School 2 56 19.18%
School 5 50 17.12%
School 1 47 16.10%
School 7 (Music Specialist School) 43 14.73%
School 4 43 14.73%
School 3 38 13.01%
School 9 9 3.08%
School 8 6 2.05%
Socio-economic 1 50 17.12%
Index 2 0 0.00%
(Based Upon 3 6 2.05%
Deciles Set by 4 52 17.81%
the Australian 5 0 0.00%
Bureau of Statistics) 6 47 16.10%
7 0 0.00%
8 56 19.18%

9
 9 43 14.73%
10 38 13.01%
Note. N = 292. Percentages may not add to 100.00% across categories within a demographic
variable due to rounding. Decile 1 = most disadvantaged; decile 10 = most advantaged

Socio-economic index. ANCOVAs showed that children from schools falling in socio-
economic decile 4 (relative deprivation) displayed significantly larger rates of singing
improvement over time on song 1 than their counterparts from schools within deciles
1, 6, or 8. Participants attending schools within socio-economic decile 10 also
demonstrated superior singing improvement rates than those from schools within
either deciles 1 or 8. Findings for song 2 were similar, with the school from decile 6
also showing outstanding results (see Figure 3).

4.00
Control
3.50
Treatment
3.00

2.50

2.00

1.50

1.00
Pre-Intervention Post-Intervention
Figure 1. Mean ratings of student singing ability according to time of
assessment, as moderated by intervention group. Error bars represent
standard errors. Comparisons were performed between the pre-intervention
and post-intervention time points within each of the control and treatment
intervention conditions. * p < .05 ** p < .01 *** p < .001.

Table 2. Descriptive Statistics for Improvement in Rated Singing Ability on Song 2

According to Age.

10
 Age (in Years) N M SE
4 1 1.00 0.00
5 73 0.41 0.09
6 84 0.76 0.07
7 62 0.68 0.09
8 4 0.25 0.25

Student survey results

Change in general attitudes towards music and singing. The analysis was restricted to
those children who had been in classes experiencing the mentoring program, as no
control group children completed the survey at the second time point. A paired
samples t-test was conducted comparing pre-intervention to post-intervention
attitude levels. However, it should be noted that very few post surveys were returned
and that post-intervention data was supplied only by schools in deciles 4 and 10.
Results showed that attitudes towards music expressed by children following
intervention (M = 2.52, SE = 0.04) were significantly more positive than those recorded
prior (M = 2.42, SE = 0.04), t(54) = 2.92, p = .005, η2 = .14 (see Figure 4). Therefore, the
intervention appears to have been successful at promoting enhanced positive
attitudes toward music.

Gender. An independent samples t-test assessing the effect of gender on pre-

intervention attitudes revealed that girls (M = 2.55, SE = 0.02) reported significantly
more positive general baseline attitudes toward music than boys (M = 2.38, SE = 0.03),
t(204) = 4.80, p < .001, η2 = .10 (see Figure 5).

11
1.20 1.20
*
1.00 1.00 *
0.80 0.80
0.60 0.60
0.40 0.40
0.20 0.20
0.00 0.00
r1

r2

r1

r2
n

n
te

te
ea

ea

ea

ea
ar

ar
Y

Y
rg

rg
de

de
in

in
K

Year Level of Student

Improvement for Song 1 Improvement for Song 2

Figure 2. Improvement in rated singing ability on Song 1 and Song 2 according

to student year level. Error bars represent standard errors. Comparisons were
conducted among all year levels. * Denotes significant results once the α = .05
critical level was adjusted via the Holm-Bonferroni technique to compensate
for multiple comparisons on overall family-wise error rate.

12
 1.40 1.40
1.20 1.20
1.00 1.00
0.80 * 0.80
0.60 ** ** 0.60 **
0.40 0.40 ***
**
0.20 0.20
0.00 0.00
1 3 4 6 8 10 1 3 4 6 8 10

Socio-Economic Index

Improvement in Rated Singing Ability on Song 1 Improvement in Rated Singing Ability on Song 2

Figure 3. Mean improvement in student singing ability on Song 1 and Song 2

based on school socioeconomic decile. Error bars represent standard errors.
Comparisons were conducted among all decile category levels. * Denotes
significant results once the α = .05 critical level was adjusted via the Holm-
Bonferroni technique to compensate for multiple comparisons on overall
family-wise error rate. Decile groups denoted by an asterisk differed
significantly from the group with a mean bar of the same colour.

However, post-mentoring survey results demonstrated a similar rate of increase in

positive attitudes toward music by both boys (M = 0.09, SE = 0.06) and girls (M = 0.11,
SE = 0.04), where an independent samples t-test upon the attitude change scores
showed no significant gender effect, t(53) = −0.33, p = .744, η2 < .01.

Age and school year level. Investigation of the developmental impact of age on general
music attitudes showed that younger age was associated with more positive overall
views concerning music, R2 = .04, F(1, 204) = 7.74, β = −.19, p = .006. However, age
was not a significant predictor of the change in attitudes to music following
intervention, R2 = .004, F(1, 53) = 0.22, β = −.06, p = .641. This suggested that it may be
more appropriate to explore this developmental effect not through use of a
continuous variable such as age, but rather through a categorical one such as year
level. A one-way independent samples ANOVA was used to inves- tigate this aspect
and showed that year level significantly affected children’s baseline general attitudes
towards music, F(2, 203) = 5.55, p = .004, η2 = .05. As shown in Figure 6, kinder-

13
garteners (M = 2.61, SE = 0.05) were found to express more favorable attitudes overall
toward music at pre-intervention than either the year 1 (M = 2.42, SE = 0.03), t(203) =
3.30, p = .001, or year 2 children (M = 2.46, SE = 0.03), t(203) = 2.42, p = .016.
General Attitudes toward Music

3.00

2.80
Scale Score (1-3)

**
2.60

2.40

2.20

2.00
Pre-Intervention Post-Intervention

Assessment Time Point

Figure 4. Mean expressed positive attitudes regarding music at each
assessment time point. Error bars represent standard errors. * p < .05 ** p
< .01 *** p < .001.
General Attitudes toward Music
Scale Score (1-3)

**

14
 3.00
2.90
2.80
2.70
2.60
2.50
2.40
2.30
2.20
2.10
2.00
Male Female

Figure 5. Average positive attitudes toward music at pre-intervention, broken

down by student gender. Error bars represent standard errors. * p < .05 ** p
< .01 *** p < .001.

Socio-economic index. An ANOVA on pre-intervention attitudes revealed that the

positivity of baseline general viewpoints regarding music varied significantly among
the socio-economic index deciles, F(5, 200) = 5.13, p < .001, η2 = .11. A series of post
hoc t-tests were employed among each possible pairing of the decile levels. As shown
in Figure 7, children who attended schools that fell within socio-economic decile 1 (M
= 2.59, SE = 0.05) reported significantly more positive general attitudes about music
than their peers from schools in either decile 6 (M = 2.31, SE = 0.05), t(200) = 4.34, p
< .001, or decile 4 (M = 2.39, SE = 0.05), t(200) = 3.11, p = .002. Further, children
attending schools within socio-economic decile level 8 (M = 2.54, SE = 0.04) also
expressed more favourable music attitudes than their counterparts in decile 6, t(200) =
3.77, p < .001. The remainder of contrasts produced non-significant results after
adjusting the α critical level for multiple comparisons, ts < 2.53, ps > .012.

15
 A second one-way independent samples ANOVA on change scores, however,
revealed that the socio-economic standing of the school did not impact significantly on
the improvement rate of positive music attitudes from pre- to post-intervention, F(1,
53) = 0.0002, p = .988, η2 < .01. However, it should be noted that post-intervention
data was supplied only by schools in deciles 4 and 10.
General Attitudes toward Music

3.00

2.80
Scale Score (1-3)

2.60

2.40

2.20

2.00
Kindergarten 1 2

Student Year Level

Figure 6. Mean general positive attitudes to music at pre-intervention

assessment, according to student year level. Error bars represent standard
errors. Planned comparisons were performed among all year levels, each at α
= .05. * p < .05 ** p < .01 *** p < .001, where the most conservative
significance level was denoted to mark group differences.
General Attitudes toward Music
Scale Score (1-3)

*
**

16
 3.00

2.80

2.60

2.40

2.20

2.00
1 4 6 8 9 10

Figure 7. Average pre-intervention music attitude scores based on

school socioeconomic decile. Error bars represent standard errors.
Comparisons were conducted among all index decile levels. *Denotes
significant results once the α = .05 critical level was adjusted via the
Holm-Bonferroni technique to compensate for multiple comparisons.
Decile groups denoted by an asterisk differed significantly from the
group with a mean bar of the same colour.

Discussion

The study investigated the nature of the impact of the mentor-supported, generalist
teacher-led music program in early childhood school settings on children’s singing skills
and attitudes to music. The findings suggest that the intervention had been successful,
with children whose teachers participated in the teacher-mentoring program
demonstrating significant improvement, regardless of gender, with children in the
control group showing no improvement over time. The positive results of this program
are in line with previous research on singing interventions (Welch et al., 2010; Welch

17
et al., 2014; Demorest et al., 2017; Runfola & Etopio, 2012). However, the findings
contradict a number of gender-focused studies that showed girls outperforming boys
in early music outcomes (Mang, 2006; Roberts, 2016; Saunders et al., 2010).
Children in year 1 produced superior rates of singing ability improvement over
time on both the easier song 1 and the more challenging song 2, in comparison to their
kindergarten and year 2 counterparts. This may imply perhaps that there may be a
developmental window at this age for music learning, although this is not particularly
evidenced in the research literature where music learning is typically reported at all
ages across childhood (Saunders et al., 2010), or no differences are reported between
different year levels (Roberts, 2016).
This age effect in favour of year 1 children might be masking a more important,
and perhaps relevant, finding that there may be a greater chance of nurturing
children’s singing development at an age before musical identities as singer/non-singer
become entrenched. Such an interpretation is in line with the research data from a
large-scale study in England involving over 11,000 children (Welch et al., 2010). Here,
nationwide singing pedagogy interventions had the greatest impact on the youngest
age groups (ages 4–8) compared with older children.
Overall, the inference from the socio-economic data is that successful singing
development can be nurtured irrespective of children’s relative levels of deprivation.
This finding is supported by recent Australian research (McPherson et al., 2015;
Osborne, McPherson, Faulkner, Davidson, & Barrett, 2016) that evaluated the impact
of music programs in low socio-economic schools. The generalist teachers of the
schools in deciles 4 and 10 had the same mentor, per- haps implying that the results
for socio-economic index for song 1 are more likely reflective of the skill and expertise
of the intervention facilitator in question, than of any socio-economic advantage or
disadvantage experienced by the schools. Similarly, results for song 2 highlight the
expertise of another mentor who was solely responsible for mentoring of all
participants falling into decile 6. This lends strength to our finding that quality music
education can produce positive music outcomes for children regardless of the socio-
economic background.
Children’s participation in the mentoring program classes increased positive
general attitudes toward music, regardless of gender, or socio-economic standing of
the school. This result is in line with previous research that highlighted the impact of
school-based music activities on children’s social and life skills (Rabinowitch, Cross, &
Burnard, 2013; Rickard et al., 2013; Rinta, Purves, Welch, Elmer, & Bissig, 2011; Welch
et al., 2014). There was an effect of gender on baseline attitudes, but not on attitude
change. The higher starting baseline for girls has been highlighted by the literature
(Roberts, 2016; Saunders et al., 2010; Sergeant & Welch, 2009). Our study

18
demonstrates that, despite this, the rate of increase in positive attitudes toward music
due to the mentoring context was equivalent for both boys and girls.
Despite the higher baseline positive attitudes toward music in younger
children, as previously flagged by research (Roberts, 2016; Saunders et al., 2010) age
did not influence the rate of increase in positive music attitudes seen following
intervention. There was an effect of year level on the pre-intervention scores, where
kindergarteners showed more positive general base- line attitudes towards music than
either their year 1 or year 2 counterparts. This may be explained by an increased
interest in other life areas introduced in the formal schooling years that may
overshadow music.
No clear pattern emerged regarding differences in attitudes towards music
with regard to the socio-economic index, and the attitudes did not change over the
mentoring period. This finding is limited because the only participants to supply
appropriate post-intervention data came from the deciles 4 and 10, thus restricting
any strong conclusions that can be drawn about the effect of educational socio-
economic standing on student music attitude change over time.

Conclusions

This study provides evidence that a mentored, generalist teacher-led music program in
early childhood education settings can positively impact young children’s singing skills
and attitudes to music. Children’s singing improved regardless of gender or socio-
economic standing of the school. Overall, the intervention was also successful at
improving general positive attitudes toward music in children, whilst noting that there
was a general bias towards reason- ably positive attitudes at baseline. Further, these
attitude improvement rates were generalizable across gender, age, and socio-
economic standing of the school and not limited to specific sub-groups.
Overall, the findings suggest that employing specialist music mentors to guide
and support generalist teachers in an intensive program can deliver positive changes in
their students’ singing skills and attitudes to music. The presence of a formal
mentoring program, as exampled by the NMTMP, can greatly enrich provision and,
arguably, make excellent use of relatively sparse specialist expertise across a broad
school population. Whatever the school context, there is a growing wealth of research
evidence to suggest that sustained successful music education can provide wider
benefits for all children, socially, emotionally, and intellectually. The results of this

19
intervention study provide further empirical evidence of the benefits accessible in
music, as well as, by implication, that are likely to accrue through music.
This study evaluated the effectiveness of a pilot music mentoring program in
formal education settings with young children. A national rollout of the scheme is now
in progress. Future evaluations of larger numbers of children across the nation will
provide more detailed and nuanced findings that will enhance music educators’
understanding of best-practice approaches to developing singing skills in children and
fostering their love of music. Future research could also focus on developing suitable
interventions for the middle primary years as, although teachers and students may
face different challenges during this period, the issues of generalist (non-specialist)
teachers’ confidence and perceptions of appropriate expertise in music education are
commonplace.

Funding
The author(s) disclosed receipt of the following financial support for the research,
authorship, and/or publication of this article: This research was funded by a grant
auspiced by the Australian Youth Orchestra.

References

ABS. (2013). 2033.0.55.001—Socio-economic Indexes for Areas (SEIFA), Data Cube

2011. Retrieved from http://www.abs.gov.au
ABS. (2014). 1269.0, Standard Australian Classification of Countries (SACC), 2011,
Version 2.3. Retrieved from http://www.abs.gov.au
Brown, E. D., Benedett, B., & Armistead, M. E. (2010). Arts enrichment and school
readiness for children at risk. Early Childhood Research Quarterly, 25, 112–124.
Costa-Giomi, E. (2004). Effects of three years of piano instruction on children’s
academic achievement, school performance and self-esteem. Psychology of
Music, 32(2), 139–152.
Costa-Giomi, E. (2012). Music instruction and children’s intellectual development: The
educational con- text of music participation. In R. MacDonald, G. Kreutz & L.
Mitchell (Eds.), Music, health and wellbeing (pp. 339–372). Oxford: Oxford
University Press.
Demorest, S., Nichols, B., & Pfordresher, P. Q. (2017). The effect of focused instruction
on young chil- dren’s singing accuracy. Psychology of Music.
doi:10.1177/0305735617713120.
DEST. (2005). National review of school music education: Augmenting the diminished.
Canberra: DEST.

20
Habibi, A., Damasio, A., Ilari, B., Veiga, R., Joshi, A. A., Leahy, ... Damasio, H. (2017).
Childhood music training induces change in micro and macroscopic brain
structure: Results from a longitudinal study. Cerebral Cortex, 8, 1–12. doi:
10.1093/cercor/bhx286.
Hall, C. (2005). Gender and boys’ singing in early childhood. British Journal of Music
Education, 22(1), 5–20.
Hallam, S. (2011). The power of music: Its impact on the intellectual, social and
personal development of children and young people. International Journal of
Music Education, 28(3), 269–289.
Hennessy, S. (2012). Improving primary teaching: Minding the gap. In G. E. McPherson
& G. F. Welch (Eds.), The Oxford handbook of music education (Vol. 2). Oxford:
Oxford University Press. doi:10.1093/ oxfordhb/9780199928019.013.0046
Johnson, C. M., & Memmott, J. E. (2006). Examination of relationships between
participation in school music programs of different quality and standardized test
results. Journal of Research in Music Education, 54(4), 293–307.
Letts, R. (2015). Tertiary pre-service courses for primary school specialist music
teachers. Music Trust of Australia. Retrieved from
http://musicinaustralia.org.au/index.php?title=Tertiary_Pre-service
Courses_for_Primary_School_Specialist_Music_Teachers
Liao, M-Y., & Davidson, J. W. (2016). The effects of gesture and movement training on
the intonation of children’s singing in vocal warm-up sessions. International
Journal of Music Education, 34(1), 4–18. Macdonald, R. A. R. (2013). Music,
health, and well-being: A review. International Journal of Qualitative Studies in
Health and Well-being, 8, 1–13.
Mang, E. (2006). The effects of age, gender and language on children’s singing
competency. British Journal of Music Education, 23(2), 161–174.
Matthews, J. S., Ponitz, C. C., & Morrison, F. J. (2009). Early gender differences in self-
regulation and academic achievement. Journal of Educational Psychology,
101(3), 689–704.
McPherson, G. E., Osborne, M. S., Barrett, M. S., Davidson, J. W., & Faulkner, R. (2015).
Motivation to study music in Australian schools: The impact of music learning,
gender and socio-economic status. Research Studies in Music Education, 37(2),
141–160.
Mecke, A-C., & Sundberg, J. (2010). Gender differences in children’s singing voices:
Acoustic analyses and results of a listening test. Journal of Acoustical Society of
America, 127(5), 3223–3231.

21
Moreno, S., Bialystok, E., Barac, R., Schellenberg, E. G., Cepeda, N. J., & Chau, T. (2011).
Short-term music training enhances verbal intelligence and executive function.
Psychological Science, 22(11), 1425–1433.
Moritz, C., Yampolsky, S., Papadelis, G., Thomson, J., & Wolf, M. (2013). Links between
early rhythm skills, musical training, and phonological awareness. Reading and
Writing, 26(5), 739–769.
Nichols, B. E. (2016a). Critical variables in singing accuracy test construction: A review
of literature. Update, 35(1), 39–46.
Nichols, B. E. (2016b). Task-based variability in children’s singing accuracy. Journal of
Research in Music Education, 64(3), 309–321.
Nichols, B. E., & Wang, S. (2016). The effect of repeated attempts and test–retest
reliability in children’s singing accuracy. Musicae Scientiae, 20(4), 551–562.
Osborne, M. S., McPherson, G. E., Faulkner, R., Davidson, J. W., & Barrett, M. S. (2016).
Exploring the academic and psychosocial impact of El Sistema-inspired music
programs within two low socio-economic schools. Music Education Research,
18(2), 156–175.
Pecina, K. (2017). Singing teachers and the development of young voices (Unpublished
master’s thesis). Macquarie University, Australia.
Portowitz, A., Lichtenstein, O., Egorova, L., & Brand, E. (2009). Underlying mechanisms
linking music education and cognitive modifiability. Research Studies in Music
Education, 31(2), 107–128.
Rabinowitch, T-C., Cross, I., & Burnard, P. (2013). Long-term musical group interaction
has a positive influence on empathy in children. Psychology of Music, 41(4), 484–
498.
Rauscher, F. H., & Hinton, S. C. (2011). Music instruction and its diverse extra-musical
benefits. Music Perception, 29(2), 215–226.
Rickard, N. S., Appelman, P., James, R., Murphy, F., Gill, A., & Bambrick, C. (2013).
Orchestrating life skills: The effect of increased school-based music classes on
children’s social competence and self-esteem. International Journal of Music
Education, 31(3), 292–309.
Rinta, T., Purves, R., Welch, G., Elmer, S. S., & Bissig, R. (2011). Connections between
children’s feelings of social inclusion and their musical backgrounds. Journal of
Social Inclusion, 2(2), 34–57.
Ritchie, L., & Williamon, A. (2011). Primary school children’s self-efficacy for music
learning. Journal of Research in Music Education, 59(2), 146–161.
Roberts, J. C. (2016). ‘Wanna race?’: Primary student preference for competitive or
non-competitive singing games. British Journal of Music Education, 33(2), 159–
174.

22
Runfola, M., & Etopio, E. (2012). Effect of music instruction on preschoolers’ music
achievement and emergent literacy achievement. Bulletin of the Council for
Research in Music Education, 192, 7–27.
Sachs, M., Kaplan, J., Der Sarkissian, A., & Habibi, A. (2017). Increased engagement of
the cognitive control network associated with music training in children during
an fMRI Stroop task. PLoS ONE, 12(10): e0187254.
Saunders, J., Varvarigou, M., & Welch, G.F. (2010). The role of singing. In S. Hallam & A.
Creech (Eds.), Music Education In the 21st Century In the United Kingdom:
Achievements, analysis and aspirations. (pp 69-84). London: Institute of
Education.
Schellenberg, E. G. (2004). Music lessons enhance IQ. Psychological Science, 15(8),
511–514.
Sergeant, D.C., & Welch, G.F. (2009). Gender differences in Long-Term-Average
Spectra of children's singing voices. Journal of Voice, 23(3), 319-336.
Southgate, D. E., & Roscigno, V. J. (2009). The impact of music on childhood and
adolescent achievement. Social Science Quarterly, 90(1), 4–20.
Welch, G.F., Himonides, E., Papageorgi, I., Saunders, J., Rinta, T., Stewart, C., Preti, C.,
Lani, J., Vraka, M., & Hill, J. (2009). The National Singing Programme for primary
schools in England: an initial baseline study. Music Education Research, 11 (1). 1-
22.
Welch, G.F., Himonides, E., Saunders, J., Papageorgi, I., Preti, C., Rinta, T., Vraka, M.,
Stephens Himonides, C., Stewart, C., Lanipekun, J., & Hill, J. (2010). Researching
the impact of the National Singing Programme ‘Sing Up’ in England: Main
findings from the first three years (2007-2010). Children’s singing development,
self-concept and sense of social inclusion. London: International Music Education
Research Centre, Institute of Education.
Welch, G. F., Himonides, E., Saunders, J., Papageorgi, I., & Sarazin, M. (2014). Singing
and social inclusion. Frontiers in Psychology, 5, 803. doi:
10.3389/fpsyg.2014.00803.
Williams, K. E., Barrett, M. S., Welch, G. F., Abad, V., & Broughton, M. (2015).
Associations between early shared music activities in the home and later child
outcomes: Findings from the longitudinal study of Australian children. Early
Childhood Research Quarterly, 31, 113–124.

23
Appendix A: Student singing assessment

School Student Initials

Year Student DOB __/ __ / __
Class Name Student Gender M F
Date Special Needs
Administered by Ethnicity

Speech-Centre Test: Ask the student to count backwards from 10 to get a sense of their speaking pitch
(expected to be around middle C (C4) and up to the E above middle C (E4). The marks on the keyboard are
for reference and indicate Middle C (hollow circle) and concert A (filled circle).

Singing Test: no starting note is given; student is asked to sing alone.

Singing Assessment Criteria (after Welch, 1998)
The words of the song appear to be the initial centre of interest rather than the melody, singing is “chant-
1 like”, employing a restricted pitch range and melodic phrases. Descending patterns tend to predominate.
2 Sung melody outline begins to follow the general contours of the target melody and its phrases. Pitch range
in singing expands.
Melodic shape and intervals are mostly accurate, but some changes in tonality may occur. Overall, the
3
number of “different” pitches is much reduced.
4 No significant pitch or rhythm errors.

24
Appendix B: Class survey (developed by Author 4)

School Student
Initials
Year Student DOB __/ __ / __
Class Name Student M F
Gender
Date Special Needs
Administered by Ethnicity
1 I feel good about myself

2 The boys in my class are

better at music than the girls
3 I like the music that I make
at school
4 I like making music

5 Singing is fun

6 I like listening to music

7 I have many friends

8 Members of my family tell

me I am good at music
9 I sing at home

10 My friends teach me songs

11 I like making music with my

friends
12 I like singing in the
playground
13 Singing is easy

14 I am the best singer in my

class
15 Singing is something that
everyone can do
16 Singing makes me feel
happy
17 I prefer to sing on my own

18 I find reading easy

19 I like going to school

20 Writing is hard

25
 21 I feel left out of things at
school
22 I am a friendly person

23 I am very happy

24 My body is healthy

26

View publication stats