Answers are grounded in this paper and cite their sources.
It is shown that the ratio of the harmonic mean of citations over the harmonic mean of publications does not lead to an acceptable impact measure for a meta-journal. This result contrasts markedly with the corresponding cases in which the arithmetic or geometric average is used. The relation between different averages and the regression of impact over publications or the regression of the opposite of impact over the opposite of publication numbers is studied in some detail. This leads to the general observation that if the regression line of y over x has a positive slope then this is not necessarily true for the regression line of 1/y over 1/x.
No takes yet. Share an insight, caveat, or question.
Egghe et al. (2013) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: