COMPARISON OF TRADITIONAL AND RESPONSE TIME ITEM RESPONSE THEORY MODELS IN THE ESTIMATION OF EXAMINEES' ABILITY
Abstract
Theoretically, Response Time (RT) Item Response Theory (/RT) models have been adjudged to be superior to their counterpart traditional IRT models in the estimation of examinees' ability, although empirical evidences that justify this claim are scarce. This study examined the comparability of the ability of examinees estimated using traditional and
RT /RT models. The study is a non-experimental research of correlational type. Multistage sampling procedure was adopted in the selection of 874 students who participated in the study The instrument used for data collection was a Computer-Based Mathematic Achievement Test (r = 0.89), The responses of the examinees to the test items and the time taken
to respond to the test items were obtained and analysed using traditional 3-parameter logistic and RT /ognormal /RT modelsas well as the paired sample t-test. Result showed that the ability estimate of examinees under traditional and RT /RT models were quite similar. However, the relative standing of the examinees ability score under RT and traditional IRT
models were different from one another. While RT IRT model was not superior to traditional RT model in the estimation of examinees ability estimate when average ability estimates of examinee is considered, the relative standing of the examinees ability were different under the two models. Recommendations were made based on the finding.
Keywords: Response time item response theory model, traditional item response theory
model, ability estimate, relative standing of ability estimates
Omonike R. Lawal & Michael Akinsola Metibemu
COMPARISON OF TRADITIONAL AND RESPONSE TIME ITEM RESPONSE THEORY MODELS IN THE ESTIMATION OF EXAMINEES' ABILITY
Abstract Theoretically, Response Time (RT) Item Response Theory (/RT) models have been adjudged to be superior to their counterpart traditional IRT models in the estimation of examinees' ability, although empirical evidences that justify this claim are scarce. This study examined the comparability of the ability of examinees estimated using traditional and RT /RT models. The study is a non-experimental research of correlational type. Multistage sampling procedure was adopted in the selection of 874 students who participated in the study The instrument used for data collection was a Computer-Based Mathematic Achievement Test (r = 0.89), The responses of the examinees to the test items and the time taken to respond to the test items were obtained and analysed using traditional 3-parameter logistic and RT /ognormal /RT modelsas well as the paired sample t-test. Result showed that the ability estimate of examinees under traditional and RT /RT models were quite similar. However, the relative standing of the examinees ability score under RT and traditional IRT models were different from one another. While RT IRT model was not superior to traditional RT model in the estimation of examinees ability estimate when average ability estimates of examinee is considered, the relative standing of the examinees ability were different under the two models. Recommendations were made based on the finding. Keywords: Response time item response theory model, traditional item response theory model, ability estimate, relative standing of ability estimates