INVARIANCE OF ITEM PARAMETERS OF POLYTOMOUS GEOGRAPHY ITEMS IN Senior SCHOOL Certificate Examination IN NIGERIA
No Thumbnail Available
Date
2018
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
UNIVERSITY OF ILORIN
Abstract
The hallmark of a test instrument is its quality. Test validity is key in ascertaining the quality of tests in all school subjects and it is critical at the School Certificate Level. Geography is examined at this level and a lot of studies have been conducted to determine the item parameters of objective tests using traditional Classical Test Theory. Therefore, it has become necessary to utilize the Item Response Theory procedure for analyzing the invariance of Item Parameters of Polytomous Items in Geography at the Senior School Certificate Examination Level in Nigeria. The objectives of the study were to examine: (i) examinees’ ability estimates; (ii) item difficulty and discrimination estimates; (iii) the extent of invariance of item difficulty estimates across male and female students; and (iv) the extent of invariance of item difficulty estimates across Geography students in rural and urban areas.
This study employed the descriptive survey design. The population comprised all senior secondary school students in Nigeria. The target population was all senior secondary school three (SSS3) students that offered Geography in their final examinations in 2016/2017 academic session. A sample of 1,546 drawn out of 32,499 students from 876 senior secondary schools in Nigeria was selected. A multi-stage sampling procedure involving stratified random sampling, simple random sampling and proportionate sampling technique were used to select the respondents. The research instruments were questions on Element of Practical and Physical Geography (WASSCE 2016 Geography Paper 3). The data collected were analyzed using descriptive statistics, Robust Maximum Likelihood Estimation and Bayesian Ability Estimation methods and to answer the research questions, while t-test statistics was employed for hypotheses testing.
The findings of the study were that:
1. the examinees’ ability estimates in Physical Geography showed that 49.35% of the examinees had performance below average, while 50.65% of the examinees had performance above average;
2. item 2B was the most difficult with difficulty index of 62.366 (mean = 3.719) ;
3. item 1B discriminated most with discrimination index of 4.364 (mean = 0.5076);
4. five out of 23 items representing 22% of the test items showed obvious variation in the difficulty estimates of the test items across gender;
5. 14 out of 23 items were more difficult for students of urban schools (Mean = -0.11, SD = 3.754) than the students of rural schools (Mean = -1.98, SD = 4.420);
6. there was a significant invariance of item parameter estimates across male and female geography students; and
7. there was a significant invariance of item parameter estimates across rural and urban geography students.
The study concluded that the items in the 2016 WASSCE Paper 3 had a large variation of item parameters based on school location (rural and urban) which implies that the test was not good for group comparison. Based on the findings, the study recommended among others that examination bodies should adopt Item Response Theory to reduce measurement problems encountered in score comparison and test equating.
Description
Keywords
INVARIANCE, ITEM PARAMETERS, POLYTOMOUS, GEOGRAPHY, Senior SCHOOL Certificate, Examination, NIGERIA