20201231172157D4978 - Psikometri 6 - 8

Course Code : Z1870
Course Name : Psychometrics
Classical Test theory and

Reliability
Session 7 - 9
TODAY’S OUTLINE
• Basic concept of reliability

• Different type of reliability
• Reliability and standard error measurement
Learning Outcome
Explain the Analyse the test
Describe what theory and or scale items
measurement is practice of critically and
and its various psychological make
aspects in testing and the comparison of
psychological important test or scale
research aspects of items
psychometry
References
Raykov, Tenko & marcoulides, Geoge A. 2011.
Introduction psychometric theory, Rouletge . New
York . ISBN 0-203-84162-X
Linda Crocker and James Algina. 2008. Introduction to

Classical and Modern Test Theory. Cengage Learning.
Ohio. ISBN:978-0-495-39591-1
Classical test theory
The Nature of Measurement
Error
Measurement Error
Systematic Random
error error
Foundations of
Classical Test Theory
1. The Classical Test Theory equation →

2. A Definition of True Score →
3. Relationships Between True Score and error Score →
Corr(T, E) = 0
Models Based on Classical
Test Theory
1. Model of Parallel Tests
Based on the CTT Equation Xk = Tk + Ek (k = 1,…,p), the

model of parallel tests asserts that all p tests (measures) X1
through Xp share the same true score; thereby, all
differences among their observed scores come from
differences in their error scores that are also assumed to be
equally varied, i.e., have the same variance. That is, the
model of parallel tests (MPT) assumes that Xk = T + Ek and
Var(Ek)=Var(Ek’)
2. Model of True Score equivalent Tests (Tau-equivalent Tests)
The assumptions made in the MPT—that the p tests measure

the same true score with the same precision (error variances)
—are often quite strong and restrictive in behavioral and
social research.
According to this model, it is required not only that the same
true score be measured in the same units of measurement
with the given set of tests but also that they all are equally
error prone (meaning, with identical error variances). Both of
these assumptions are limiting substantially the range of
applicability of the MPT in empirical work, making it rather
difficult for substantive scholars to construct measures that
fulfill it. It is therefore desirable to have available alternatives
that do not make these strong assumptions.
3. Model of Congeneric Tests
The assumption of the same true score being evaluated with

the same units of measurement by a set of given tests, which
underlies both the MPT and TSEM, is rather strong and
restrictive in the behavioural and social sciences. It requires
not only that the same latent continuum be evaluated by all
tests (measures) but in addition that they all should be based
on the same units of measurement.
In behavioural and social research, however, units of
measurement are quite often lacking particular meaning,
arbitrary or perhaps even irrelevant, in addition to being
specific to different tests and therefore likely to be distinct
across tests. This assumption of equality of measurement units
does indeed strike one as fairly strong, because it requires what
seems to be in general far too much from these arbitrary units
that cannot be typically meaningfully interpreted.
Reliability
Basic concept of reliability
Reliability of a test is a criterion of test quality relating to the

accuracy of psychological measurements. The higher the
reliability of a test, relatively the freer it would be of
measurement errors.
The concept of reliability underlines the computation of
error of measurement of a single score, whereby we can
predict the range of fluctuations likely to occur in a single
individual score as a result of irrelevant chance factors.
Different type of reliability
There are fives methods to measure the reliability of a test.

These are,
a) test–retest method,
b) method of parallel form,
c) split-half reliability,
d) method of rational equivalence and
e) Cronbach’s Alpha.
• Test–retest method
Two trial • Method of parallel form
• Split-half reliability
Single trial • Method of rational equivalence
• Cronbach alpha
Test–Retest Method
The most frequently used method to find the reliability of a test

is by repeating the same test on a second occasion.
The reliability coefficient (r) in this case would be the
correlation between the score obtained by the same person on
two administrations of the test.
1st Test 2nd Test

The formula used to find the test–retest reliability is the
Pearson product–moment formula.
Where,
N = total numbers of observations,
X = Scores in first data set,
Y = Scores in second data set.
Split-half Reliability
The advantage that this method has over the test–retest

method is that only testing is needed. This technique is also
better than the parallel form method to find reliability
because only one test is required.
A reliability coefficient of this type is called a coefficient of

internal consistency.
In this method, the test is scored for the single testing to get
two halves, so that variation brought about by difference
between the two testing situations is eliminated.
The two halves of the test can be made by counting the

number of odd-numbered items answered correctly as the
other half.
1 2 3 4 5 6 1 3 5 ∑ 2 4 6 ∑
A 0 0 0 0 0 0 0 0 0 0 0 0 0 0
B 0 0 0 0 1 0 0 0 1 1 0 0 0 0
C 1 0 1 1 1 0 1 1 1 3 0 1 0 1
D 1 1 1 1 1 1 1 1 1 3 1 1 1 3
E 1 1 1 1 1 1 1 1 1 3 1 1 1 3
F 0 0 1 0 0 0 0 1 0 1 0 0 0 0
G 0 0 1 1 1 0 0 1 1 2 0 1 0 1
H 1 1 1 1 1 0 1 1 1 3 1 1 0 2
I 0 0 0 1 0 0 0 0 0 0 0 1 0 1
J 0 1 0 1 0 1 0 0 0 0 1 1 1 3
these halves will be correlated with the help of the Pearson

product–moment formula,
The reliability depends upon the test length. When we score
as two halves, we, in fact, cut the length of the original test to
half.
Therefore, the reliability which we have calculated is

equivalent of one for a test of half of the size of our original
test. Thus, we make the correlation of test length to get the
reliability of the total original test
For this Spearman–Brown formula is used for doubling the
length of the test. The formula is,
Method of Rational Equivalence
The coefficient of internal consistency could also be obtained

with the help of Kuder–Richardson formula number 20
To compute reliability with the help of Kuder–Richardson

formula number 20, the following procedure is used.
First, write the first column in a worksheet showing the
number of items.
The second, column should give the difficulty value (p) of each
item obtained during item analysis.
The third, column is given as q where q = 1 – p.
The fourth, column is taken as (p) (q). This column is the product
of column two and column three.
The Kuder–Richardson formula number 20 is,
Where N is the number of items on the test, σ2 is the variance

of the test and σ2t is the standard deviation, M is the mean of
the test score
Cronbach Alpha
• The Kuder–Richardson formula is applicable to find the

internal consistency of tests whose items are scored as
right or wrong, or according to some other all or none
system.
• Some tests, however, may have multiple choice items.
• On a personality inventory, however, there are more than
two response categories.
For such tests, a generalised formula has been derived known
as coefficient alpha (Cronbach 1951).
In this formula, the value of Σpq is replaced by Σσt2 which is the

sum of variance of item scores.
The procedure is to find the variance of all individuals scores for

each item and then to add these variances across all items.
• The formula is,
where
n = number of items in test
σt2 = variance of test scores
σi2 = item ith variance
Σ σi2 = sum total of items variance
Reliability and standard error
measurement
THE formula to compute sem is
Where
σE = standard error measurement
σX = standard deviation
ρXX’ = reliability coeffecient

20201231172157D4978 - Psikometri 6 - 8

Uploaded by

Copyright:

Available Formats

20201231172157D4978 - Psikometri 6 - 8

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

20201231172157D4978 - Psikometri 6 - 8

Uploaded by

Copyright:

Available Formats

Course Code : Z1870

Course Name : Psychometrics

Classical Test theory and

• Basic concept of reliability

Linda Crocker and James Algina. 2008. Introduction to

1. The Classical Test Theory equation →

1. Model of Parallel Tests

Based on the CTT Equation Xk = Tk + Ek (k = 1,…,p), the

The assumptions made in the MPT—that the p tests measure

The assumption of the same true score being evaluated with

Reliability of a test is a criterion of test quality relating to the

There are ﬁves methods to measure the reliability of a test.

The most frequently used method to ﬁnd the reliability of a test

1st Test 2nd Test

The advantage that this method has over the test–retest

A reliability coefﬁcient of this type is called a coefﬁcient of

The two halves of the test can be made by counting the

these halves will be correlated with the help of the Pearson

Therefore, the reliability which we have calculated is

The coefﬁcient of internal consistency could also be obtained

To compute reliability with the help of Kuder–Richardson

Where N is the number of items on the test, σ2 is the variance

• The Kuder–Richardson formula is applicable to ﬁnd the

In this formula, the value of Σpq is replaced by Σσt2 which is the

The procedure is to ﬁnd the variance of all individuals scores for

THE formula to compute sem is

You might also like