Taking Note of Our Biases: How Language Patterns Reveal Bias Underlying the Use of Office Discipline Referrals in Exclusionary Discipline

Abstract

The comments teachers write when sending students to the office have the potential to increase our understanding of how bias may contribute to longstanding racial disparities in school discipline. However, large-scale analysis of open text has traditionally had a prohibitive cost. Through natural language processing techniques, we examined over 3.5 million office discipline records from national samples of more than 4,000 schools for whether teachers’ linguistic patterns differed when describing incidents depending on the race/ethnicity and gender of the students. Results of such analyses consistently showed that teachers wrote longer descriptions and included more negative emotion when disciplining Black compared to White students, especially for Black girls. In conjunction with psychology of language theory, the patterns suggest that teachers may perceive and process student behavior differently depending on student identities. Implications of the findings and potential for research on naturally occurring language data in education are discussed.

“All the Makings of an Eventual NFL Starter:” Racial/Ethnic Disparities in a Decade of National Football League Draft Profiles as Revealed by Natural Language Processing

“…Principles of the stereotype content model have been evaluated verbally in several studies to demonstrate how people of color are communicated about compared to White people (for examples in different societal domains, see Markowitz et al, 2023;Nicolas et al, 2021;Voigt et al, 2017). Critically, several papers have also observed disparities in how people communicate about athletes of different demographic backgrounds.…”

“…Recall, using computational and natural language processing techniques, prior work has applied this words-as-attention model to bias in different settings, including medicine (Markowitz, 2022a), policing (Voigt et al, 2017), education (Markowitz et al, 2023), and sport (Fu et al, 2016). In sport, specifically, journalists who interviewed players tended to focus on different aspects of the match depending on player gender.…”

“…For example, the sentence "This player is a hard worker" contains 6 words and LIWC identifies the following categories, including but not limited to: impersonal pronouns (this; 16.67% of the total word count), achievement words (e.g., hard worker; 33.33%), and articles (e.g., a; 16.67%). All categories were drawn from the standard LIWC-22 dictionary, and this tool has been applied to a range of bias evaluations in other settings (Markowitz, 2022a;Markowitz et al, 2023;Voigt et al, 2017).…”