Drivers of academic performance in a Brazilian university under a government-restructuring program

摘要

The search for correlates of scientific production is an important step toward the formulation of decision-making guidelines on academic and funding policy under a competitive system with continuously reduced budgets. Our goal here is to identify drivers of the scientific production of researchers working at the “Universidade Federal de Goiás” (UFG), a medium-to-large public Brazilian University, focusing on the effects of teaching load and supervision of graduate and undergraduate students on scientific production of faculty members. We analyzed data for 1487 faculty members of UFG, including the total number of papers published between 2011–2013, a weighted-index of scientific production and the number of publications in high-ranked journals (according to a Brazilian system of journal ranking in different areas). These variables were regressed on gender, teaching load at undergraduate and graduate levels, number of supervised undergraduate, Master and Doctoral students, self-declared amount of time dedicated to research and outreach, year of doctoral graduation, year of hiring and the scientific production before doctoral graduation. Several regression models were used to model scientific production, including ordinary least-square regression and Hurdle negative binomial models. Although there are some differences among research areas, the most important explanatory variable was the publication record of the researcher before doctoral graduation, reinforcing the role of a solid academic formation in terms of research experience. Undergraduate and graduate teaching loads were negatively and positively correlated with scientific production, respectively. However, the strength of the relationship was much higher for the second than for the first relationship. These correlates of scientific production provide guidelines for policy and management in universities, including criteria for balancing research and teaching loads, awarding fellowships and research grants, designing new policy for future hiring and creation of new graduate programs.