Rethinking how we measure gender associations

This study introduces a novel norming method for measuring the gender associations of words that avoids binary Likert scales that present male and female as opposite endpoints. Here, participants provide ratings on two separate scales for a word's likelihood of referring to a woman or a man, respectively. This approach allows for a more nuanced measurement, independently capturing bias toward each gender rather than assuming they are inversely related. Results from the two-scale method are strongly correlated with traditionally used binary scale ratings, which suggests that this new method reliably captures gender bias of words without reinforcing a binary continuum. In addition to nouns, the present study extends gender norming to adjectives, which have received comparatively little attention in prior experimental work. We find that adjectives exhibit gender associations similar to those observed for nouns, which provides evidence that gendered expectations extend beyond nouns in English

Binary data · Binary number · Gender bias · Likert scale · Noun · Evolutionary Psychology and Human Behavior · Gender Studies in Language · Neurobiology of Language and Bilingualism