Psychometrics properties of math questions of university entrance examination 2012/2013 academic year of mathematics and physics participants in Iran

Subject Areas :

Morteza Hosseini ¹ , Ali Moghaddamzadeh ² , Mostafa Rasteghar Aghah ³

1 - PhD Student in Psychometrics, Allameh Tabatabaei University , Department of Educational and Psychological Measurement (Psychometrics), Tehran, Iran
2 - Assistant Professor, University of Tehran, Faculty of Psychology and Educational Sciences, Tehran, Iran.
3 - PhD Student in Psychometrics , Allameh Tabatabaei University ,Department of Educational and Psychological Measurement (Psychometrics), Tehran, Iran.

Received: 2022-04-15 Accepted : 2022-08-29 Published : 2025-01-23

Keywords: &quot, High-stakes tests&quot, , &quot, CTT&quot, , &quot, IRT&quot, , &quot, measurement&quot, , &quot, Psychometric properties&quot, .,

Abstract :

Abstract: The purpose of this study was to examine psychometric properties of mathematics items of university entrance examination 2012/2013 academic year of mathematics and physics students in Iran. The method of this study was descriptive and applied. To do so, 1000 math university participants were randomly selected out of all of participants. For data analysis, SPSS-19 and Bilog-mg was used. All items analyzed based on CTT and IRT. Then Item difficulty and item parameters were calculated in terms of both IRT and CTT. CTT results showed that math examination was so hard with good reliability and validity. As well, some items had low discrimination index. Three of items had negative discrimination index that had major problems based on CTT. The assumptions and the goodness of fit analysis were checked for IRT. The results showed that this is a Unidimensional test with good Local independence. The items that had fit with none of models are: 1-2-3-4-6-12-14-15-17-19-21-26-29-38-40-41-43-44-45-51-54. Also, the most appropriate model which fit the data was achieved by 2pl logistic IRT model. IR analysis based on 2pl showed that none of items had negative discrimination index. Also, 17 items had not good discrimination index. Item difficulty of items in IRT such as CTT showed that math questions were so hard. These results were showed test information curve.