Monday, 11 July 2016

The h-index: meritocracy or spurious reductionism?

Niall McCrae, King’s College London


A recent JAN editorial has put a few ‘noses out of joint’. Editor-in-chief Roger Watson, far from the metropolitan elites in the cool climes of Hull, has emulated the irreverent poet Philip Larkin, whose verbal bombs from that smoky port were as loud as those of the Luftwaffe’s blitz in the Second World War.

Surveying the publication activity of the nursing professoriate, Watson et al. (2016) found several cases of chairs of nursing departments whose output was no better than that of a novice research worker. According to the Scopus database, some had h-index scores of 1; this could amount to a single publication, cited once. Watson et al.’s exposé has drawn an angry response, but as the Lancaster bomb crews once said: the greater the flak, the closer you are to the target.

An eloquent riposte by Gary Rolfe (2016) in the Journal of Clinical Nursing showed Watson et al.’s findings to be flawed, partly due to reliance on a database of dubious comprehensiveness. Scopus misses some nursing journals, authors’ details are frequently incorrect, and a more fundamental problem in the blunt collation of data is in the loss to female scholars whose names changed on marriage.

Accepting some of the criticisms, Watson would prefer Web of Science for a future survey, and the ORCiD number should be used to prevent disadvantage to women. Perhaps someone could take this on, but the results will not be any more impressive across the board. My h-index in Scopus lingers at a modest 8; in Web of Science it is a paltry 6. These numbers are frustratingly low for a writer with two books, numerous book chapters and over 50 published papers. By contrast, Google Scholar would give me a reassuring score of 12, being more inclusive in journal coverage and in types of citation.

No metric will satisfy all, but the h-index is an objective measurement not merely of the number of publications but the author’s impact. A paper in The Lancet will contribute much more to the score than one published in the East Yorkshire Archives of Health & Postmodern Poetry (impact factor unavailable).

A nursing professoriate emerged later in the UK than in the USA, and it seems that confidence in the academic prowess of the discipline is not as well developed here as in other countries. Deans of nursing faculties or chairs of nursing departments have a vital role in contributing to the science and knowledge of healthcare, and publication is the vehicle for changing and improving practice. Rolfe suggested that professors be rated by a mystical p-index, for which no criteria were offered. The h-index may seem crude, but at least it is objective, whereas a nebulous notion of ‘influence’ would be highly subjective – and could perpetuate an 'old boys club'.

Indeed, Rolfe’s argument, while persuasive on many points, has a degree of the reactionary tone of the establishment. Professors of nursing are not suited to an ivory towers mentality, and they should be able to face criticism from the likes of Watson et al., who will keep us on our toes. Let’s not shoot the messenger.


References

Rolfe G (2016) Professorial Leadership and the h-index: the rights and wrongs of academic nursing Journal of Clinical Nursing DOI: 10.1111/jocn.13428

Watson R, McDonagh R, Thompson DR (2016) h-indices: an update on the performance of professors in nursing in the UK Journal of Advanced Nursing DOI: 10.1111/jan.12924


No comments:

Post a Comment