What Age Is in a Name?

Sasha Johfre

doi:10.15195/v7.a15

Abstract

Social scientists often describe fictional people in survey stimuli using first names. However, which name a researcher chooses may elicit nonrandom impressions, which could confound results. Although past research has examined how names signal race and class, very little has examined whether names signal age, which is a highly salient status characteristic involved in person construal. I test the perceived demographics of 228 American names. I find that most strongly signal age, with older-sounding names much more likely to be perceived as white than as black. Furthermore, participants' perceptions of the age of a name poorly match with the true average birth year of people with that name, suggesting that researchers cannot simply use birth records as a proxy for perceived age. To assist researchers in name selection, I provide a set of candidate names that strongly signal a matrix of combined age, race, and gender categories.

Highlights

Social scientists often describe fictional people in survey stimuli using first names
I used birth record data from the Social Security Administration (SSA) on frequencies of babies born to every name each year from 1880 to 2018.2 I selected 90 names with a high probability of being age-typed by first searching for names whose mean and mode birth year was within the same decade (1930–1939, 1940–1949, 1950– 1959, 1960–1969, 1970–1979, 1980–1989, 1990–1999, 2000–2009, 2010–2018), and selecting the 10 names from each decade that had the smallest standard deviation in number of babies per year to get highly time-specific names
Race, and parental education perceptions were not associated with the magnitude of the error. These results indicate that truly older names and low-popularity names are associated with age perceptions that differ from what would be expected from looking at birth records

Summary

Introduction

Social scientists often describe fictional people in survey stimuli using first names. The signaling power of names suggests that the particular name used in a survey stimulus may cause research participants to have systematic impressions about the fictional person being discussed. Some scholars interested in differences based on a target’s race have used only name difference to signal the race category membership of a fictional person (Bertrand and Mullainathan 2004; Correll et al 2007; Gaddis 2015) Their name selection is based on a significant body of work showing how names signal race and class (Barlow and Lahey 2018; Crabtree and Chykina 2018; Gaddis 2017b, 2017a; Goldstein and Stecklov 2016).

Methods

Results

Conclusion