Beyond frequencies: a cross-national network analysis of basic skills in Latin American higher education

How are basic skills represented across higher education systems, and what explains their distribution? We address this through a bipartite network analysis linking 3,155 postgraduate programs from 481 universities in nine Latin American countries to ten O*NET basic skills. Descriptive analyses reveal a highly stable cross-national skill hierarchy, with scientific reasoning consistently occupying the most prominent position. However, inferential network models show that this apparent stability conceals important structural processes. Skill–program ties are associated with informational, structural, institutional, and contextual factors, with heterogeneity across skills accounting for the largest share of variation in the network. Scientific reasoning remains dominant after controlling for these factors, whereas writing and mathematical reasoning remain comparatively peripheral. The findings demonstrate that frequencies alone provide an incomplete account of skill representation and that inferential network models offer a richer explanation of how educational priorities are structured across higher education systems.