Alternative metrics for measuring the quality of articles and journals.

Richard Smith

doi:10.3332/ecancer.2013.ed18

Abstract

Alternative metrics for measuring the quality of articles and journals.

Highlights

I’ve fretted about this problem since 1991 when I became the editor of the BMJ and chief executive of the BMJ Publishing Group responsible for about another 15 journals. In those days academics in Britain and editors of specialist journals were not familiar with impact factors, but once they began to be used to measuring the performance of academic departments—and, most important of all, allocating money to them— academics began to be obsessed by impact factors
Editors too became obsessed and would weep when their journal impact factors fell by one decimal point
The New England Journal of Medicine had an impact factor in 2012 of 53.298 but most journals have impact factors under 1

Summary

Introduction

I’ve fretted about this problem since 1991 when I became the editor of the BMJ and chief executive of the BMJ Publishing Group responsible for about another 15 journals. The impact factor of a journal for, say, 2012 (and it is measured annually) is citations in 2012 to articles published in 2010-2011 divided by number of “citable” articles published in 2010-2011. When it comes to journals, the impact factor has some validity because the articles in the journal being highly cited must be some sort of measure of impact.

Results

Conclusion