VALIDITY & RELIABILITY

Mrs Deborah Esan

Introduction
 Reliability and validity are two important and essential
concepts that relate to each potential instrument that
the investigator is considering.

 They have specific definitions in research and can be

easily confused
 Validity refers to whether the instrument actually
measures what it is supposed to measure.

 For Example: Does the instrument measure the concept

being examined?
 Reliability refers to whether the tool conveys consistent
and reproducible data, for example, from one participant
to another or from one point in time to another.
VALIDITY
Validity is the degree to which a tool measures
what it is supposed to measure (Mateo, 1999).
Assures that a test is measuring what it is
intended to measure
There are three types of validity;
content,
 criterion-related, and
construct.

Validity relies first and foremost on reliability

Validity Contd.
 Content validity relates to an instrument’s adequacy
in covering all concepts pertaining to the phenomena
being studied.

 For example;If the purpose of the tool is to learn

whether the patient is anxious before taking an exami-
nation, the questions should include a list of behaviors
that anxious peo-ple report when they are
experiencing anxiety.
.
 It is important that an instrument be reviewed for
content by persons who possess characteristics and
experiences similar to those of the participants in a
study.
Validity Contd
 Criterion-related validity, which can be either predictive or
con-current, measures the extent to which a tool is related
to other criteria.
 Predictive validity is the adequacy of the tool to estimate
the individual performance or behavior in the future.
 It deals with association with a subsequent event thought
to be an outcome –pre-maturity and death

 For example, if a tool is developed to measure clinical

competence of nurses, persons who respond to the tool can
be followed over time to see if this score correlates with
other measures of competence, such as performance ap-
praisals, commendations, or other indications of
competence.
Validity Contd
 . If results indicate that there is a high coefficient
correlation, it means that the clinical competence
scale can be used to predict future performance
Appraisals.
 Concurrent validity is the ability of a tool to
compare the
 respondent’s status at a given time to a criterion
(Mateo, 1999).
 For ex-ample, when a patient is asked to complete a
questionnaire to determine the presence of anxiety,
results of the test can be compared to the same
patient’s ratings on an established measure of
anxiety administered at the same time.
Validity Contd
 Construct validity is concerned with the ability of
the instrument to adequately measure the
underlying concept (Mateo, 1999).
 With this type of validity, the researcher’s concern
relates to whether the scores represent the degree
to which a person possesses a trait.
 Construct validity deals with how measure relates
to other measures- attitude to HIV patients
measured by reaction to PLWHA

 Since construct validity is a judgment based on a

number of studies, it takes time to es-tablish this
type of validity.
Types of validity
Face validity
Indicates that an instrument appears to test
what it is supposed to
Content validity
Indicates that the items that make up an
instrument adequately sample the universe
of content that defines the variable being
measured
Most useful with questionnaires and
inventories
Criterion-related validity
Indicates that the outcomes of one
instrument, the target test, can be used as a
substitute measure for an established
reference standard criterion test
Can be tested as concurrent or predictive
validity
Concurrent validity establishes validity when
2 measures are taken at relatively the same
time, most often used when the target test is
considered more efficient than the gold
standard, and can be used instead of the
gold standard
Predictive validity establishes that the
Construct validity
Establishes the ability of an instrument to
measure an abstract construct and the
degree to which the instrument reflects the
theoretical components of the construct
RELIABILTY
The reliability of a tool is the degree of
consistency in scores achieved by sub-jects
across repeated measurements.
It is the extent to which a measurement is
consistent and free from error
Also known as reproducibility or
dependability.
Consistency of measurement or its stability
The comparison is usually reported as a
reliability coefficient.
 The reliability coefficient is determined by the
proportion of true variability (attributed to true
differences among re-spondents) to the total
obtained variability (attributed to the result of true
 differences among respondents and differences
related to other factors).
 Reliability coefficients normally range between 0
and 1.00; the higher
 the value, the greater the reliability.
 In general, coefficients greater than 0.70 are
considered appropriate.
 However, in some circumstances this will vary.
Acceptable reliability
Poor reliability : less than 0.5
Moderate reliability: 0.5 to 0.75
Good reliability: greater than 0.75
Researchers may be able to tolerate lower
reliability for measurements used for description
Measurements used for decision making or
diagnosis need to be higher, perhaps at least
0.9 to ensure valid interpretation of findings
The researcher takes a chance that data across
repeated administra-tions will not be consistent
when instruments with reliability estimates of
0.60 or lower are used (Mateo, 1999).
Three aspects should be considered when
determining the reliabil-ity of instruments:
stability,
equivalence, and
homogeneity
The stability of a tool refers to its ability
to consistently measure the phenomenon
being studied.

This is de-termined through test-retest

reliability.
Test-retest reliability

The tool is administered to the same

persons on two separate occasions. Scores
of the two sets of data
are then compared, and the correlation is
derived. The recommended interval
between testing times is 2 to 4 weeks
(Burns & Grove, 2005; Mateo, 1999)
 Equivalence should be determined when two versions
of the same tool are used to measure a concept
(alternative forms) or when two or more persons are
asked to rate the same event or the behavior of an-
other person (interrater reliability; Mateo, 1999).
 In alternative-form reliability, two versions of the same
instrument are developed and ad-ministered. The
scores obtained from the two tools should be similar.
 It is helpful for the researcher to know whether a
published instrument has alternate forms; when there
are, a decision must be made about which
 form to use.
Establishing interrater reliability is important when
two or more observers are used for collecting data.
The homogeneity of an instrument is determined
most commonly by calculating a Cronbach’s alpha
coefficient.
This test is found in a number of statistical
packages.
This test is a way of determining whether each
item on an instrument measures the same thing.
Internal consistency reliability estimates are also
calculated by using the Kuder-Richardson formula.
Types of reliability
Test-retest reliability
Rater reliability
Alternate forms reliability
Internal consistency
Test retest reliability
 Sample of individuals is subjected to the
identical test on two separate occasions,
keeping all testing conditions as constant
as possible
 The test-retest reliability coefficient is
reported
 The coefficient is indicative of reliability in
situations where raters are not involved
e.g. self-report survey instruments and
physical and physiological measures with
mechanical or digital readouts
Rater reliability
Many clinical measurements require that a
human observer, or rater, be part of the
measurement system
Could be
Intrarater reliability (stability of the data
recorded by one individual across two or
more trials)
Interrater relaibility (variation between two
or more raters who measure the same group
of subjects)
Alternate forms reliability
Many measuring instruments exist in 2 or
more versions, called equivalent or parallel
forms
Often used as an alternative to test-retest
reliability when the nature of the test is
such that subjects are likely to recall their
responses to test items.
Internal consistency
Extent to which items measure items
measure aspects of the same characteristic
Most common approach involves looking at
the correlation among all items in a scale
Split half reliability
THANK YOU