Supplementary Information for Reciprocity and success in academic careers

Weihua Li,Tomaso Aste,Fabio Caccioli,Giacomo Livan
2019-01-01
Abstract:In S. 1, we detail the results of a few linear regression models for authors whose career started between 1970 and 1990. The dependent variable is the average number of citations per paper for models M1-2, while the dependent variable for models M3-4 is the average total number of citations per author. In both cases those quantities are computed at the 20th career year for each author. The average number of publications per author is used as an independent variable in all models. Moreover, the average number of citations at the 10th career year and the average number of citations at the 10th career year are used as independent variables for models M1-2 and M34, respectively. In all columns, variables such that p < .05 are highlighted with one asterisk, while variables such that p < .01 are highlighted with two asterisks. Excess reciprocity is used as an additional independent variable for models M2 and M4. As it can be seen from the Rsquared values, the introduction of excess reciprocity as an additional independent variable does not provide any meaningful statistical improvement to the models. Yet, let us remark that in model M2, i.e., the only one where excess reciprocity is statistically significant, the relationship between
What problem does this paper attempt to address?