Researcher capacity estimation based on the Q model: a generalized linear mixed model perspective

Abstract Chance models of scientific creative productivity allow estimation researcher capacity. One prominent such model is the Q in which impact a scholarly work modeled as multiplicative function capacity and potential (i.e., luck) parameter. Previous estimated based on an approximation In this work, however, I outline how can be within framework generalized linear mixed models. This way estimates (and all other parameters model) are readily available obtained by standard statistical software packages. Usage further allows comparing different distributional assumptions calculation reliability parameter capacity). illustrated for large dataset multidisciplinary scientists ( N = 20,296). The Poisson was found to have negligibly better predictive accuracy than original normal model. Reliability revealed excellent with conditional being mostly acceptable ranges. depended heavily number publications scientist increasing papers. future limitations context thoroughly discussed.