Reconsidering the Reference Category

Abstract

Social scientists often present modeling results from categorical explanatory variables, such as gender, race, and marital status, as coefficients representing contrasts to a “reference” group. Although choosing the reference category may seem arbitrary, the authors argue that it is an intrinsically meaningful act that affects the interpretability of results. Reference category selection foregrounds some contrasts over others. Also, selecting a culturally dominant group as the reference can subtly reify the notion that dominant groups are the most “normal.” The authors find that three of four recently published tables in Demography and American Sociological Review that include race or gender explanatory variables use dominant groups (i.e., male or white) as the reference group. Furthermore, the tables rarely state what the reference is: only half of tables with race variables and one-fifth of tables with gender variables explicitly specify the reference category; the rest leave it up to the reader to check the methods section or simply guess. As an alternative to this apparently standard practice, the authors suggest guidelines for intentionally and responsibly choosing a reference category. The authors then discuss alternative ways to convey results from categorical explanatory variables that avoid the problems of reference categories entirely.

US county-level prevalence and spatial distribution of optimal birth outcomes 2018–2019

“…We defined counties achieving greater racial birth equity as those where the total population EBS optimal birth rate was higher than the national 75th percentile and the absolute difference between maternal racial/ethnic categories (non-Hispanic Black vs. non-Hispanic White and Hispanic vs. non-Hispanic White) was smaller than the national 25th percentile difference. While using White as a reference group is not always appropriate—due to the group’s increasing heterogeneity and the possibility of further entrenching existing racial inequities 17 —this analysis uses it only in a descriptive manner to estimate the disparity—that is the numeric difference between two populations—of optimal births between Whites, who comprise the majority of the US population, and Blacks and Hispanics, two populations within whom heightened risks exist for adverse outcomes. In order to visualize this measure, maps were created in ArcMap 10.7 showing the estimated health-equity status of each county.…”

“Husband, father, coward, killer”: The discursive reproduction of racial inequality in media accounts of mass shooters

“…Descriptors for shooters racialized as Hispanic are almost never statistically significant in this analysis, meaning that descriptive 4 We rely on Black as the reference category here for two reasons. One is that models using culturally dominant groups as the reference categories sometimes operate in ways that implicitly frame culturally dominant groups as "normal" -statistically examining how various Others vary from what is framed as the norm (e.g., Johfre and Freese, 2021). Here, we select Black as the underabundance of descriptors used in news articles on Black shooters helps us to better appreciate the media treatments groups racialized in other ways receive in comparison to Black shooters.…”

Connecting QuantCrit to Gifted Education Research: An Introduction

“…Educators' perceptions of how race/ethnicity and racism operate in our society have major implications specifically on how minoritized racial/ethnic groups are treated in our education systems and we must continue to challenge the systems perpetuating injustices, including academic scholarship. Researchers should also choose variables, models, and groups based on theoretical framing and use alternatives such as contrasts when possible (Johfre & Freese, 2021).…”