Fig. 3

Exemplary graphical model test with 95% confidence regions and gender as splitting criterion. Item numbering does not correspond to the final item numbering. The estimated item parameters for males are plotted on the x-axis, and the estimated item parameters for females are plotted on the y-axis. Item parameter estimates are surrounded by 95% confidence regions. Significant deviation from the diagonal indicates an item’s subgroup heterogeneity. For items 3 and 7, parameter estimations differed markedly between males and females. Item 3, which addressed the deceleration of a car that is moving with constant speed, was easier to solve for males. Item 7, however, which addressed the interaction between the Earth and the Moon, was easier to solve for females