Development of A Job Stress Scale-20 by using Item Response Theory: study of reliability and validation

Objective: The purpose of this study was to develop a new scale to measure a job stress, and examine its some psychometric properties. Methods: The study was carried out in three different phases and studies. In the first phase of the study was created item pool and applied. A total of 2073 people, 898 women and 1175 men ranging in age from 18 to 65 participated in this study. In the second phase of the study selected items were administered and the total number of participants were 2565, 1452 women and 1113 men, ranging in age from 18 to 69. In the third phase of the study final form of 20 items were administered and the total number of participants were 615 (295 women and 320 men) ranging in age from 19 to 60. Out of these participants, 108 people (48 women, 60 men) took part in the test-retest study. Further, in phase third for criterion related validity all participants completed the A Stress Scale, the Perceived Stress Scale, the Burnout Measure, the Hospital Anxiety and Depression Scale, the Zung Depression Scale, the Personal Well-Being Index, the Warwick-Edinburgh Mental Well-Being Scale and the Oxford Happiness Questionnaire. Results: For the analysis of the items in the first phase, item analysis with Item Response Theory were used, item discrimination and item difficulty coefficients were calculated, and 24 items were selected. In the second phase, 20 items were selected with same way, and the latest version of the form was named as A Job Stress Scale-20. Developed new scale explained 38.07% of the total variance by single factor and confirmatory factor analysis results showed model fit. Cronbach's alpha internal consistency reliability coefficient for the whole scale was computed 0.91 in the second and third phase. Correlation coefficient among the test-retest administration with two weeks of intervals was determined 0.88 in the third phase. In addition, it was observed that correlations ranged between -0.40 and 0.65 for the other scales with new job stress scale. Discussion: Although the need for further validity studies, initial results shows that the A Job Stress Scale-20 has met the basic psychometric require-ments. [Anadolu Psikiyatri Derg 2020; 21(0.100): 14-22]