Comparative Analysis of Number Correct and Item Pattern Scoring Methods for Waec and Neco Chemistry Items Among Senior Secondary Students in Osun State

Volume 4 Issue 1 (January 2022)
Temitope Babatimehin Adamu Chidubem Deborah Adeoye Oluseyi Peter
Pages: 32-39 Download Count : 1095 View Count: 1565 DOI Number 10.17220/tuara.2022.01.04 Facebook Share on Google+ Save to Zotero Save to Mendeley

Abstract:

The study determined the unidimensionality using Item Response Theory (IRT) also compared the number correct and item pattern scoring methods of Senior School Certificate Examination (SSCE) examinee estimated scores from WAEC and NECO chemistry items in Osun State, Nigeria. This is with the view of providing empirical explanation on the appropriateness of decisions made statistically from examinee scores using classical test and item response theory approaches. The study adopted a descriptive survey design. A sample of 1,105 students was randomly selected from a population of 36,182 students who sat for the 2017/2018 SSCE chemistry paper 1 in Osun State. An adopted version of June/July 2015 NECO (Type 1) and May/June 2015 WAEC (Type 2) SSCE Chemistry (Objective) Paper 1 titled Chemistry Achievement Test (CAT) were utilized for data collection. Data collected were analyzed using STEU implemented in DIMTEST mean, standard deviation, paired sampled t-test, MIRT and equate IRT packages of R language. The results showed that the traits measured by the two groups of items were significantly different from one another (multidimensionality). Also, results showed that IRT (item pattern) scoring method has significant effect on the examinees scores in WCI (t = 13.06, p < 0.05) and NCI (t = 7.667, p < 0.05) than CTT (number correct) scoring method. In conclusion, the study established that there is no statistical equivalence between IRT (item pattern) and CTT (number correct) scoring method on WAEC and NECO chemistry items in Osun State.

Keywords

  • Comparative analysis
  • Item response theory; Classical test theory
  • number correct and item pattern scoring methods
Download PDF