Using Multiple Regression in the Hurwicz Method to Determine the Cut Score of Tests as an alternative to Angoff and Bookmark Methods
Using Multiple Regression in the Hurwicz Method to Determine the Cut Score of Tests as an Alternative to Angoff and Bookmark Methods
Subject Areas : Instruction
maryam parsaiean
1
,
Ebrahim Khodaie
2
,
balal Izanloo
3
,
Keyvan Salehi
4
,
Sima Naghizadeh
5
1 - Department of Curriculum development & amp; instruction methods. Faculty of Psychology and Education Tehran University
2 - Department of Curriculum Development & Instruction Methods, Faculty of Psychology and Education University of Tehran
3 - Associate Professor, University of Kharazmi, Faculty of Psychology and Education , Tehran,
4 - Associate Professor, Faculty of Psychology and Education, University of Tehran, Tehran, ds. Faculty of Psychology and Education Tehran University
5 - Assistant Professor, National Organization of Educational Testing (NOET), Tehran, Iran,Email
Keywords: Cut score, Hurwicz method, multiple regression, Angoff method, bookmark method, cross-validation, Tolimo test,
Abstract :
The aim of the current research is to determine the cutoff score of the Tolimo test for doctoral candidates applying for study opportunities abroad by applying multiple regression in the Hurwicz method, which is a decision-making approach. The statistical population of this research includes the candidates of one course of Tolimo exam, which were 461 people. The research method of this study is based on a quantitative approach, and in terms of its purpose, it is part of applied research, and in terms of analysis, it is part of secondary analysis studies, and the scores of all 461 candidates in each of the grammar, reading comprehension, listening, and writing sections are used to determine the cutoff score. becomes Based on the results of the study and considering the small amount of classification error, if the total score is used, the average value of the Hurwicz index is the cutoff score of 481, and if the linear regression model is significant, using the results of the Hurwicz index in multiple regression, the cutoff score is 491. It can be used as a cutoff score. The proposed method significantly improves the accuracy of determining the cutoff score compared to Angoff and Bookmark methods. According to the findings obtained from cross-validation and error values, the cutoff score obtained from the combined strategy leads to more accurate results with zero classification error. Keywords: Cut score, Hurwicz method, multiple regression, Angoff method, bookmark method, cross-validation, Tolimo test
