تقصي الثبات لمقياس استراتيجيات التعلم المنظم ذاتيا في ضوء النسب المفقودة وطرق تعويضها
This study aims to investigate the stability coefficients (alpha for Cronbach and omega for McDonald) for the Self-Regulated Organizational Learning Strategies Scale using the methods of maximum likelihood and multiple imputation. This exploration was conducted in scenarios where there were no missing values and in cases where missing values were present at different proportions, with subsequent treatment using both maximum expectation and multiple imputation techniques. To achieve the study's objectives, a scale of Self-Regulated Organizational Learning Strategies consisting of 77 items was administered to a randomly selected sample of 980 undergraduate students from various disciplines at Yarmouk University. Data were subjected to complete random missingness at rates of 5%, 15%, and 30%. The missing values were then inputted using both maximum expectation and multiple imputation techniques. Stability coefficients (alpha for Cronbach and omega for McDonald) were calculated for the scale in each research scenario. The study's results indicated that the values of alpha and omega coefficients obtained using the maximum expectation method were higher than those obtained with the multiple imputation method, across various rates of missingness (5%, 15%, 30%). The study recommends paying more attention to missing data in the design and analysis of studies, understanding the reasons behind data loss, avoiding arbitrary selection of methods for handling missing data, using multiple imputation methods for data comparison, and considering the use of omega for McDonald's coefficient as an alternative to Cronbach's alpha. (Published abstract)