Exploring individual differences in the development of sociolinguistic bias: A large-scale study

• Accent-based in-group preferences emerge by age 3. • Preferences increase non-linearly across development, with sharp growth between ages 4.5 and 5.5. • Exposure to other languages reduces children’s accent-based preferences. • Girls show stronger accent-based biases than boys. Children tend to favor in-group members over out-group members, but how do these preferences develop across childhood, and what role does intergroup contact (exposure and/or contact with out-group members) play? Here, we focus on the developmental trajectory of accent-based preferences, and whether exposure to other languages and accents reduces these biases. In two experiments, children completed a friendship preference task, where they were asked to choose who they preferred as a friend between locally accented English speakers (in-group) and foreign-accented English speakers (out-group). In Experiment 1 ( N = 194), which tested 4-year-old children (ages 3.75–5.20 years) with at least 70% exposure to English, we found a preference for in-group speakers that increased with age, as well as an effect of English (but not accent) exposure, with children exposed to multiple languages showing weaker preferences. In Experiment 2 ( N = 483), which included a broader age range (3.5–7 years) and greater variability in English exposure (≥50%), we found a nonlinear effect of age, with preferences for local speakers emerging by age 3 and increasing most sharply around ages 4–5. Across the full sample, English exposure did not predict preferences, but in children with ≥70% English exposure, greater exposure was associated with stronger in-group preferences, replicating Experiment 1. Unexpectedly, we also found an effect of gender, with girls exhibiting stronger biases than boys. These findings indicate that accent-based biases emerge early and increase non-linearly with age, are stronger in girls, and that exposure to multiple languages may reduce these biases.