Article

Is it suitable to use the same categorization in rating scales when applied to students with distinctive levels of achievement?

---

Abstract

This study examined the efficiency of using the same rating scale categories in measuring affective constructs for students with distinctive levels of achievement. Data used in this study came from the Trends in Mathematics and Science Study (TIMSS) 2011, as a case, on the three scales that were designed to measure eighth graders' attitudes towards science. Data from the four higher and the four lower science performing countries were analyzed using Rasch model. Results revealed that the use of a four-point rating scale appears to be appropriate for some of the higher performing countries; however, it was not appropriate for the lower performing countries. In addition, category functioning and distances between threshold estimates differed by whether the country was a higher or a lower performing country; distances between threshold estimates were too close for the lower performing countries as compared to the higher performing countries. The findings of the current study question the utility of using these scales for a cross-national sample and deducing results concerning the samples’ agreeability to the construct of attitudes towards science.

Keywords: Rasch analysis, large-scale assessments, rating scale optimization

How to Cite:

Akour, M. M., Hammouri, H., Sabah, S. & Alomari, H., (2021) “Is it suitable to use the same categorization in rating scales when applied to students with distinctive levels of achievement?”, Practical Assessment, Research, and Evaluation 26(1): 18. doi: https://doi.org/10.7275/21246182