Abstract

OVALLE, Claudia Patricia. Differential Item Functioning Conditional to School Modality in the M ath PSU Test. Rev. Fac. Cienc. Tecnol. [online]. 2022, n.51, pp.169-186.  Epub Oct 18, 2022. ISSN 0121-3814.  https://doi.org/10.17227/ted.num51-11574.

This study establishes if the items of the PSU of Mathematics have a bias or DIF, being, the minority group, the students of technical modality. The method includes DIF analysis with the MH statistic (Mantel Haenszel' s X 2) and for the differences in "difficulty" and "guessing" between groups (test takers from academic vs. technical high schools) the parameters of the items were calculated with the 2pl and 3pl models. The results indicate that there is a differential functioning in all the forms of the mathematics test of the PSU test and although there are no items with critical bias or category "C", minimum or category "A" bias persists. In the 3pl model it is confirmed that although the guessing parameter is higher for the items answered by technical students, the difference in the difficulty parameters is bigger (compared to technical students). The differences range within all contents in all subtests of the mathematics psu test (algebra, geometry, numbers and probability) impacting the TP group.

Keywords : differential Item Functioning; Mantel Haenszel; Item analysis; PSU test.

        · abstract in Spanish | Portuguese     · text in Spanish     · Spanish ( pdf )