Quality Assessment of the Academic Freedom Index: Strengths, Weaknesses, and How Best to Use It

This article reviews the data quality of the first systematic global measurement of academic freedom, the Academic Freedom Index (AFI), by using a data quality assessment approach proposed by McMann et al. (2022). By analyzing three distinct components of data quality (content validity, the data generation process, and convergent validity), we examine the specific strengths and potential shortcomings of the AFI. The findings indicate that the AFI does well in terms of its theoretical embeddedness (within some conceptual limits), of the transparent data generation process, and the handling of expert assessments, as well as of its temporal and spatial coverage. A critical assessment of the level of disagreement between expert coders further shows that there are few systematic predictors, providing no evidence for problematic biases among AFI coders. Overall, we conclude that the data quality of the AFI is comparatively high but that it could be further increased by recruiting even more experts and thereby enhancing the Bayesian IRT model’s performance.