Dataset for: Of paper, panels, and pedagogy: A multi-sample analysis to establish measure-ment invariance of the Academic Risk-Taking Scale across questionnaire modes and academic disciplines

Academic risk-taking refers to students’ engagement under uncertainty about the correctness of their contributions. It reflects a learning strategy of accepting potential errors to deepen understanding and is associated with higher learning success. The Academic Risk-Taking Scale measures such behaviors, but its validity depends on demonstrating measurement invariance across relevant groups. This study examined measurement invariance across administration modes, recruitment strategies, and academic disciplines using MG-CFA on data from N = 1,259 students. Strict measurement invariance was established for recruitment mode and academic disciplines, and partial measurement invariance, from the scalar level onward, for administration mode. This level of measurement invariance permitted meaningful latent mean comparisons, revealing group differences across administration modes, recruitment strategies, and academic disciplines. These results indicate that the scale can be reliably applied across these groups, supporting its use in diverse research contexts and providing a foundation for further investigations into patterns of academic risk-taking.