Page 59 - Contributed Paper Session (CPS) - Volume 4

CPS 2126 Dr. Rajkumari Sanatombi Devi et.al
Below 0.20 Very difficult

0.20 0.50 Good
0.50 – 0.80 Best
Above 0.80 Very easy
Discrimination power (DI) defined by Blood and Budd (1972) as the ability of
an item on the basis of which the discrimination is made between superiors
and inferiors was used in assessing the discriminated power of the test items
in the present study. The formula for calculating DI was given as

DI = ( Ru - Rl)/ 0.5 N, where
DI = discrimination index
N = total no. of correct responses
Ru = the number of students in the upper 27% who responded correctly
Rl = the number of students in the lower 27% who responded correctly
Ebel and Frisbie, (1986) rule of thumb for determining the quality of items with
respect to their discrimination index was used in the present study. It was given
as

Cut off point of Discrimination index (DI) Quality of items
Above 0.40 Excellent

0.30 to 0.39 Good
0.20 to 0.29 Mediocre
Below 0.20 Poor
0.0 No discrimination
Negative value Worst

2. Justification of the study
An achievement test is one of the important aspects of teaching and
learning process which involves both the teacher and the learner. The basic
purpose of administering the multiple choice questions/items is to measure
whether the students are able to understand the content of the subject matter
of a particular subject. One of the important characteristics ofa good tests are
the reliability and validity of the test items. It is very difficult to understand
whether the test items are reliable and valid by simply looking at the options
given in a test. A reliable test items is free from bias that there is no chance of
influencing unsystematic errors in measurement even when the test are
administered to others groups in different condition by different
administrators. On the other hand, validity of a test is influences both the
unsystematic and systematic error of measurement. Hence, a test may be
reliable without being valid, but a test can’t valid without being reliable.

48 | I S I W S C 2 0 1 9