EFFECT OF ITEM PRE-KNOWLEDGE ON LOCAL ITEM DEPENDENCE AND DISTRACTER EFFICIENCY AMONG ECONOMICS STUDENTS IN OGUN STATE, NIGERIA

Abstract

This study investigated the effect of item pre-knowledge on occurrence of Local Item Dependence and Distracter efficiency. The study adopted the quasi-experimental research design. The Sample consisted of 1500 Economics students selected through multi-stage sampling technique. The research instrument used was Economics Achievement Test (EAT) which contained 60 multiple-choice items with five options scored dichotomously and adapted from 2015 NECO SSCE Economics Paper III. Fifty percent of the items were pre-known (experimental items-Form A) while the other 50% were secured items (control items-Form B). Data collected was analysed using Yen Q3 statistic, frequency counts and percentages. The results showed that the 5 pairs of items comprising 6 pre-known items and 1 secured item exhibited local item dependence for Form A of the EAT while only 1 pair of item comprising 1 pre-known and 1 secured item exhibited local item dependence for Form B . However, the 6 pre-known items in the Form A did not exhibit local item dependence when they were secured in the Form B. Also, amongst the items with Non-Functional Distractors in form A, 12 (40%) were pre-known and 6(20%) were secured. With form B, only I item (42, 3.33%) was pre-known and 3 items (item 12, 16 and 26; 10%) were secured. It was recommended among others that feasible approach to eliminate the occurrence of cheating through any form especially pre-knowledge be considered as well as items with inefficient distractors be assessed and reviewed for any further usage.