Disproportionality in Classroom Management: Measuring Implicit Bias By Coding Variation in Observable Behaviors between Classrooms

Introduction: Research has demonstrated that classroom sanctioning may not be applied in a systematic or in a unbiased manner perhaps due to implicit biases (American Psychological Association, 2016; Okonofua & Eberhardt, 2015). Students of color are sanctioned disproportionately, with higher likelihood of receiving a referral to the principal’s office, suspension or expulsion, as compared to their White counterparts (Howard, 2008). Prior to students being sent out of the classroom, however, we should consider precipitating events from which we can infer how educators biases may appear in more subtle ways. Today, there are few coding schemes that examine and measure the subtle enactments of implicit bias as revealed by classroom management. Therefore, the purpose of this study was to develop a coding variation that measures implicit bias through observable behaviors in classrooms.

Methods: Conducting secondary data analysis of classroom videos from the Measures of Effective Teaching project, a team of seven graduate students, including the PI, analyzed 66 video classrooms to develop a coding scheme. Of the 66 classrooms, 66% of teachers were White, 20% were Black, 14% were Latinx, and 1% identified as Other. Classrooms were balanced between high (>60%), medium (40-60%), and low (<40%) proportions of students of color: 37%-high; 32%-medium, 32%-low. Our goal was to determine the feasibility of documenting discrete behaviors with different implications for students’ experiences. We accomplished this by seeking a holistic understanding of classroom enactment of bias through meticulous field notes while viewing this purposively selected set of classroom videos. We met regularly to discuss our descriptive notes (e.g., triggers, events, and responses) that unfolded in the classroom. Guided by our field notes, and existing literature, we developed a multidimensional observational coding strategy.

Results: We developed an event-level set of behavioral codes to capture the proportion of specific classroom management behaviors: (1) Positive Reinforcement; (2) Negative Reinforcement; and (3) Problematic Classroom Management.

Conclusions: Our coding scheme codifies observable behaviors from teachers and what they look like and, in turn, sheds light on operational teacher practices to first measure implicit bias so as to effectively disrupt it. Implications for use of this coding scheme to support educator-focused intervention efforts are discussed (e.g., behavioral feedback; teacher training; professional development).