A multidimensional Rasch analysis of gender differences in PISA mathematics

Area of Science:

• Educational Psychology
• Psychometrics
• Gender Studies

Background:

• Historically, male advantage in U.S. standardized math tests has been noted since the 1970s.
• Girls often match boys in classroom math performance but lag on standardized assessments.
• This disparity warrants further investigation into specific mathematical domains.

Purpose of the Study:

• To compare the mathematics performance of 15-year-old males and females in the United States using PISA 2003 data.
• To identify specific math domains where gender differences in performance are most pronounced.
• To explore potential explanations for observed gender gaps in standardized testing.

Main Methods:

• Utilized data from the Programme for International Student Assessment (PISA) 2003 mathematics assessment.
• Employed a multidimensional Rasch model for accurate item calibration and ability estimation.
• Analyzed performance across four distinct mathematical domains: Space and Shape, Change and Relationships, Quantity, and Uncertainty.

Main Results:

• Small, consistent gender performance differences were observed, favoring boys, with effect sizes below .20.
• The "Space and Shape" domain exhibited the most significant gender gap.
• The "Quantity" domain showed the smallest gender difference, potentially linked to classroom familiarity.

Conclusions:

• While small, gender disparities in U.S. standardized math test performance persist.
• The "Space and Shape" domain appears particularly susceptible to gender-based performance differences.
• Familiarity with mathematical content through classroom practice may mitigate gender gaps in specific areas like "Quantity".