Page 63 - Contributed Paper Session (CPS) - Volume 4

CPS 2126 Dr. Rajkumari Sanatombi Devi et.al
Two items each were having 0 and negative discriminating power in the study.
These 4 items need to be completely removed in the construction of test items.
In a study of MCQs having 5 options, the mean score was 27.31 ± 5.75
(maximum 50 marks). Mean Dif I-value and DI were 54.14 ± 17.48 and 0.356
± 0.17, respectively. Seventy eight per cent items were of recommended
difficulty with mean 51.44 ± 11.11. Sixty two per cent items had excellent DI
(0.465 ± 0.083) (Hingorjo & Jaleel, 2012). The mean Dif I and DI were higher in
their study as compared with the present study but the variability in both Dif
I and DI were higher in the present study. In the present study, the relationship
between Dif I and DI was not in linear. It was in curvilinear form. Pearson
correlation coefficient “r” showed a significant positive correlation between
these two indices and the strength of correlation was strong (r = 0.51, P =
0.001). However, a wide scattering of dots were observed between the test
items which indicate the guessing practice done by the students. Suruchi and
Rana (2014) found that out of 120, one very difficult and seven very easy test
items were rejected for the final draft of achievement test based on the
difficulty index. Fourteen items were found to be good, 2 items needed
improvement and 6 items which fell below 0.20 of discriminating index were
rejected. One item was found to have zero discriminating power and none of
the test item showed negative discrimination. However, in their study no wide
scattering data points was observed which was in contrast of the present
study. Thus, 9 items were recommended to reject for the final draft of
achievement test. In their study, the relationship between Dif I and DI was a
moderate negatively correlated (r = - 0.3711). Mitra et a.l, (2009) reported that
difficulty and discrimination indices are reciprocally related and poor
correlation was observed in their study. Kheyami et al., (2018) found a
significant dome-shaped correlation between DIF I and DI (r = 0.162; P =
0.010), with the highest DIs occurring in the acceptable DIF I range and
decreasing for DIF Is in the difficult range which is similar with the finding of
the present study.

6. Conclusion
The finding of the study signifies about the importance of item analysis in
the selection of a reliable and validity test items in a multiple-choice types
questions. Based on the difficulty index of the test items, 31 items that
comprised of 14 good and 17 best items were retained, and 4 difficulties and
3 very easy items were rejected. Discriminating power of the test items
indicates that 10 excellent, 6 good and 11 mediocre items were retained, and
7 poor items need to be revised due to the poor discriminating power of the
test items. The reasons of the poor discriminating power may be due to the
confusing meaning of the options used test items. Sixteen items can be
considered as ideal test items based on 10 excellent and 6 good items. Two

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