The skewness of scientific productivity

摘要: Abstract This paper exploits a unique 2003–2011 large dataset, indexed by Thomson Reuters, consisting of 17.2 million disambiguated authors classified into 30 broad scientific fields, as well the 48.2 articles resulting from multiplying strategy in which any article co-authored two or more persons is wholly assigned many times necessary to each them. The dataset characterized proportion who have their oeuvre several fields. We measure individual productivity ways that are uncorrelated: number per person and mean citation period. analyze shape types distributions field using size- scale-independent indicators. To assess skewness we use robust index skewness, Characteristic Scores Scales approach. For inequality, coefficient variation. In field, study samples: entire population, what call “successful authors”, namely, subset scientists whose above average. main result that, spite wide differences production practices across very similar parallelism results for population whole successful authors, when measured person, reveals fractal nature this case. These essentially maintained fractionally