الخصائص السيكومترية والأداء التفاضلي المنتظم وغير المنتظم لمقياس الأمن النفسي باستخدام نموذج سلم التقدير لدى طلبة كلية التربية بالرستاق

The current research aimed to identify the psychometric properties available for the Psychological Security Scale using the rating scale model and to determine the extent of Uniform and Nonuniform Differential items functioning estimation using the Mantel-Haenszel generalized method and logistic regression for the Psychological Security Scale according to the student's gender and the academic year among students of the College of Education in Rustaq. The research sample consisted of 220 male and female students from the College of Education at the University of Technology and Applied Sciences, Rustaq during the second semester of the academic year 2022-2023. To achieve the research objectives, the researchers used the research tool (the psychological security scale) and applied it after verifying the hypotheses of the item response theory and the conditions of model usage. The descriptive-analytical approach was used to detect the Differential items functioning of the Psychological Security Scale items, using the statistical analysis software STATA v. 14. The research findings are as follows: Difficulty estimates scale items ranged from (-1.738: 5.287) using the rating scale model, and all of them fell within the acceptable range (-2.95: 2.95), except for items 5, 6, 14, and 32. Some of the squared chi-square values for the scale items were not statistically significant. The scale's reliability coefficient was 0.897, and the Akaike information criterion (AIC) value was 13770.72, while the Bayesian information criterion (BIC) value was 13957.37, indicating that the scale possesses good psychometric properties. Additionally, there were no statistically significant differences between male and female students of Differential items functioning estimation for the scale items using the Mantel-Haenszel generalized method, except for item 8, which showed statistically significant Differential items functioning at a significance level of 0.05 in favor of male students. Item 23 showed statistically significant Differential items functioning at a significance level of 0.01 in favor of female students, and item 28 showed statistically significant differential performance at a significance level of 0.05 in favor of female students. Furthermore, there were no statistically significant differences between male and female students in the regular Differential items functioning estimation for the Psychological Security Scale items using logistic regression, except for items 23 and 26. There were no statistically significant differences between male and female students in the irregular Differential items functioning estimation for the Psychological Security Scale items using logistic regression, except for items 14, 20, 30, 31, 40, 43, and 44. Moreover, there were no statistically significant differences between first-year and fourth-year students in Differential items functioning estimation for the scale items using the Mantel-Haenszel generalized method, except for items 38 and 42, which showed statistically significant Differential items functioning at a significance level of 0.05 in favor of fourth-year students. There were no statistically significant differences between first-year and fourth-year students in the Uniform Differential items functioning estimation for the Psychological Security Scale items using logistic regression, except for items 21 and 34. There were no statistically significant differences between male students and female students in Nonuniform Differential items functioning estimation for the Psychological Security Scale items using logistic regression, except for item 8.