Scaling the divergent correlates test for (Mednick) according to the item response theory among female secondary school students in Makkah Al-Mukarramah

Abstract

The current study aimed to grade the Mindeck divergent correlations test items to measure creative thinking ability by using a three-parameter model, and it also aimed to calculate the individuals' estimates corresponding to each probable raw score on the scale, to achieve the study objectives, the researcher followed the descriptive approach steps, and the study relied on the developed divergent correlations test items (Al Noodle and Zare’a, 2020). The test, in its initial form, consists of (46) items, which was applied on a stratified and random cluster sample consisted of (1330) female secondary school students in Makkah Al-Mukarramah city. The statistical analysis program (Bilog Mg3) was used.
The results were as following: It was possible to grade the divergent correlations test items using the three-parameter model on a single linear grading scale with only one zero and a known reliable measurement unit, which is the logit, which could be converted to the Minf unit, the final test items number was (22) items after deleting the inappropriate items for the objective measurement foundations, which was (24) items. The individuals' estimates corresponding to each probable raw score on the scale was calculated The final measurement had the required validity and reliability of the scale. In light of results, the researcher recommended the importance of providing the necessary capabilities to use the three-way model in analyzing tests models, providing the necessary computer programs, and training those in charge of the evaluation process on those programs. It also suggested conducting more studies and research related to the construction and grading of the divergent correlations test using the four-parameter model