نوع مقاله : مقاله پژوهشی
نویسندگان
1 دانشجوی دکتری سنجش و اندازهگیری دانشگاه علامه طباطبایی و عضو هیات علمی دانشگاه امام حسین (ع)
2 استاد دانشکده روانشناسی و علوم تربیتی دانشگاه علامه طباطبایی
3 استادیار دانشکده روانشناسی و علوم تربیتی دانشگاه علامه طباطبایی
4 دانشیار دانشکده روانشناسی و علوم تربیتی دانشگاه علامه طباطبایی
چکیده
کلیدواژهها
عنوان مقاله [English]
نویسندگان [English]
New theories of measurement, focusing on the characteristics of people and questions, have opened a new horizon for optimizing tests as much as possible in the direction of merit selection among candidates, especially applicants to enter military organizations. Therefore, the present research was conducted with the aim of optimizing the questions of recruitment tests of a military organization. The research method was a causal-comparative description and the statistical population was all the participants in the 1402 recruitment exam of that organization (12,000 people), and a random sample of 2,500 people was selected. The research tool was four-choice tests of mathematics, biology and literature. The findings showed that the assumption of unidimensionality was rejected in all three tests. The math and literature tests were three-dimensional and the biology test was two-dimensional. All three tests had a better fit with the compensatory model. Regarding the fit of the questions of each test, it was found that some questions were fit with the compensatory model, some with the non-compensatory model, and some with both models. Some of the questions did not fit with any of the two models. The need to observe employment justice in the selection of candidates to enter military organizations requires that the powerful techniques of item-response theory are used in the design of exam questions, and in accordance with the sensitivity of job positions, compensatory or non-compensatory multidimensional models are used in the design of exam questions.
کلیدواژهها [English]