Abstract:

Assessment at the Secondary Education level in Nigeria is taking a dimension where sophisticated instruments are being used to determine which item will be included in an item bank. The issue of pre-test or trial testing to develop the item bank is gaining grounds. It has been observed that assessment bodies in Nigeria conduct item analysis of trial-tested data to determine parameters such as difficulty, discrimination and guessing tendency. However, assessment bodies do not usually conduct post examination analyses. The purpose of this study, therefore, was to determine the post-test parameters of 60-item Mathematics test using the dichotomous and polytomous approaches. Each of the items had five options.

To carry out this study, the sixty – item Mathematics achievement test was subjected to calibration using the principles of Item Response Theory (IRT). The sample size was 16,000 – school learners who sat for the senior secondary school certificate examination in Osun State of Nigeria.

For the dichotomous scoring only one option was scored correct while all other options were scored zero. However, for polytomous scoring, each of the test options measured a particular value between 5 and 1. The most correct option carried a value of 5 while the least correct carried a value of 1.

The response of the 16,000 candidates were subjected to calibration to determine the item parameters at the dichotomous and polytomous approaches. It was found that the use of X-calibre 4.2 to estimate the item parameters of the post-test was a useful tool for assessment bodies.

Keywords:

Pre-test, ​​difficulty, ​​discrimination, ​​​​​guessing, ​​post-test, ​​dichotomous, ​​ ​​​Polytomous, ​X-calibre.