Page 71 - DEDU504_EDUCATIONAL_MEASUREMENT_AND_EVALUATION_ENGLISH

Unit 6 : Validity – Types , Methods and Usability

Notes
S 2 T = True Variance

S 2 E = Error Variance
True variance too can be further divided into two parts : (a) valid variance and (b) reliable but irrelevant
variance; that is :
2
S 2 T = S + S 2 1
V
Where, = True Variance
= Relevant or Valid Variance
= Irrelevant Variance
Thus we can say that S 2 X = S + 2 V S + 2 1 S 2 W
It can be understood more clearly by the following diagram :

S v 2 S I 2 S E 2

S X 2

It is clear from the above formula that when error variance is lower, that is, a test is more reliable,
then the ratio of valid variance increases, but because the remaining systematic variance can be
relevant or irrelevant, so a lower number of errors does not mean that its validity will be higher.
This is reason that it is said : “Reliability is a necessary, but not a sufficient condition for validity.”

6.1.2 Criterion for Validity
An independent criterion is needed for determining validity. In mental tests, the measurable elements
cannot be seen in their concrete forms, they can be perceived only through the medium in the form
of activities. Electricity cannot be seen, but it can be known by a lighted bulb or running machine.
In the same manner, intelligence cannot be seen, but it can be found out by the activities of an
individual. If an individual lacks intelligence, he cannot understand or perform even small activities;
and an intelligence individual can perform even the most difficult tasks. Therefore, in order to
establish validity of a test, some such activity will have to be found out from the external environment
in which the measurable element is manifested.

6.2 Types of Validity
According to the National Committee on Test Standards, validity of a test can generally by done
by the following three types of validity :
1. Content or curricular validity,
2. Criterion-related validity, and
3. Construct validity.
1. Content or Curricular Validity : By content validity of a test we mean each item in the test
should be a sample of that knowledge and performance for which it has been constructed.
Content validity may be defined as the extent to which a test measures a representative sample
of the subject matter and the behavioural changes under consideration.
According to Anastasi, “Content validity involves essentially the systematic examination of
the test content to determine whether it covers a representative sample of the behaviour

LOVELY PROFESSIONAL UNIVERSITY 65