Gender-Biased Perceptions Fuel Early Mathematics Gender Gap

• In Event: Gender and Mathematics: An Interdisciplinary Symposium Examining Classroom Influences, Achievement, and Affect

Abstract

Evidence from state assessments suggests that boys and girls score similarly in mathematics (Hyde et al., 2008). Yet, recent evidence from the Early Childhood Longitudinal Study, Kindergarten Class of 1998-99 (ECLS-K) suggests that although boys and girls enter kindergarten with similar mathematics scores on average (d=.00), a small gap (d=.24) favoring boys emerges by third grade (Robinson & Lubienski, 2011). Interestingly, during this K-3 period when the boys’ mathematics advantage emerges, teachers rate the math skills of girls higher than would be suggested by the tests (Fryer & Levitt, 2010; Robinson & Lubienski, 2011). Given the gender disparity in mathematics-related careers, this may seem promising news if teachers’ greater confidence in girls helps to level the playing field for future generations of women.

However, these prior estimates of teachers’ supposed female bias do not account for the better classroom behavior of girls and thus may confuse good behavior with high achievement. In this study, we disentangle these factors.

ECLS-K provides a unique opportunity to examine teacher ratings and the effects of these ratings on achievement. Data were collected in kindergarten and grades 1, 3, and 5. In each wave of data collection, U.S. D.O.E.-trained employees administered a psychometrically valid mathematics test and each student’s teacher was asked to rate the student’s mathematics proficiency. Importantly, teachers and students were never informed of the student’s score on the mathematics test.

Contrary to recent suggestions that teachers hold more positive assessments of girls’ mathematics performance, we find that teachers rate the mathematics skills of girls lower when they perform and behave similarly to boys. That is, after accounting for mathematics-achievement histories, behavior, approaches to learning, race, age, socioeconomic status, and even looking at boys and girls with the same teachers (using teacher fixed effects models), girls’ skills are rated to be more than one-tenth of a standard deviation lower than their boy counterparts. This pattern is consistent throughout elementary school. Even when accounting for current mathematics achievement, girls are still rated lower. Of note, there is no evidence of similar ratings disadvantage for black or Hispanic students; and there is no evidence that girls are rated lower or higher in reading after accounting for these factors. This suggests that the teacher underrating phenomenon is unique to girls and mathematics performance.

This tendency for teachers to underrate girls’ mathematics performance relative to boys who perform and behave similarly appears to have substantial effects on the development of gender achievement gaps in elementary school. Overall, the results suggest that if teachers’ perceptions of boys and girls were not biased against girls, then girls could lose substantially less ground in mathematics achievement in each period examined (up to 74% of the growth in the gender gap was explained by teachers’ bias in our models). As Hyde’s research (2005, 2008) demonstrates, boys and girls can achieve similarly in mathematics; thus, raising teachers’ awareness to their own unwarranted gender biases may help ensure that boys and girls have equal opportunities to reach their potential in mathematics.

Authors

• Joseph P. Robinson, University of Illinois at Urbana-Champaign

• Sarah Theule Lubienski, University of Illinois at Urbana-Champaign

• Yasemin Copur-Gencturk, University of Southern California