Graphing the Second Sex: Evidence for a Male-First Preference for Data Displays in Studies of Health-Related Quality of Life From 2003 to 2024
Author(s)
Manalastas E1, Angdembe A2, Hegarty P3
1Visible Analytics Ltd, Sheffield, YOR, UK, 2Visible Analytics Ltd, Oxford, OXF, UK, 3The Open University, Milton Keynes, BKM, UK
Presentation Documents
OBJECTIVES: Previous analysis of scientific publications in medicine and psychology has documented a simple but widespread bias in the reporting of sex differences: Graphs and tables present data representing males first, ahead of data representing females. Male-first displays perpetuate gender stereotypes, including the androcentric beliefs that men are somehow the default category of humans and that women are the ‘effect to be explained’. This research aims to document whether, and to what extent, a male-first display preference is also present in the literature on health-related quality of life (HRQoL).
METHODS: We conducted a content analysis of all articles that report on sex differences published in a leading HRQoL journal between 2003-2024. Data on graphs, tables, and study characteristics was extracted. We examined graphs and tables for sex order preference using coding procedures established in previous research on data displays.
RESULTS: We identified 233 papers featuring 503 tables and 65 graphs representing data on sex differences. Authors consistently presented male data ahead of female data in tables of baseline characteristics (61%; N=192), subgroup comparisons (72%; N=172), and regression models (82%; N=22). Likewise, majority of graphs displayed data on males first (77%; N=65). This asymmetric pattern was observed regardless of first-author gender, author country, publication year, or sex composition in samples. Exploratory analysis also suggested a third, novel manifestation of a male-first preference in data handling: regression models were more likely to arbitrarily set ‘male’ as the reference category (60%; N=74), even when women made up more than half of the sample.
CONCLUSIONS: Replicating and extending previous findings, we observed a widespread asymmetric pattern in data displays used in HRQoL studies: Tables and graphs present data on males first, ahead of data on females. This is evidence of a previously undocumented form of implicit bias in HRQoL research practice.
Conference/Value in Health Info
Value in Health, Volume 27, Issue 12, S2 (December 2024)
Code
MSR13
Topic
Health Policy & Regulatory, Methodological & Statistical Research, Organizational Practices, Study Approaches
Topic Subcategory
Best Research Practices, Health Disparities & Equity, Literature Review & Synthesis, PRO & Related Methods
Disease
No Additional Disease & Conditions/Specialized Treatment Areas