Gender guessing games: diverse methods of gendering in bibliometrics from a classification and an ethical perspective

Abstract

Purpose As gendering methods become more common in bibliometric studies, scrutinizing and developing current methodology for assigning gender is essential. This study analyses bibliometric studies' descriptions of how they gender, from a classification and ethical perspective. Design/methodology/approach The empirical material comprises 599 bibliometric studies using diverse methods for gendering authors. The methods are categorized and analysed using classification theory, where prototype theory plays a central role. Both the development and the relative growth/decline over time are described. Findings Gendering names was the most common method (240 studies), with photos (69 studies) slightly rising and linguistic cues (68 studies) declining somewhat. The classification is almost exclusively binary (women or men); only seven studies include gender diverse categories. The classification perspective reveals how several current methods are questionable both in validity and ethically. When it comes to gender diverse persons, “the paradox of exposure” needs to be considered. The discussion concludes with a proposal whereby names may be used in a specified manner, though surveys are preferable, as self-identified gender (or the option to decline answering) yields more valid and ethically sound studies. Originality/value This study offers a novel approach to understanding gendering methodology. Through the lens of classification theories – both the mutually exclusive approach (classical Greek classification) and the graded approach of prototype theory – several problems with current methodology are highlighted.