Bias and teacher perceptions

Research has shown that teacher expectations frequently influence student outcomes. American University, The Institute for the Study of Labor in Germany, and Johns Hopkins University recently collaborated on a study to determine if teachers’ perceptions of their students’ future educational attainment could be correlated with their ethnicity or gender. In other words, would teachers predict brighter futures for students who shared their race or gender than for students of other races and genders?

Researchers examined data from the Educational Longitudinal Study (ELS) of 2002, conducted by the National Center for Education Statistics, which followed 16,810 US tenth graders (age 15-16). The ELS contained predictions from each student’s maths and English teacher about how far they expected them to go in school.

No correlations were found for factors such as their grades in ninth grade, socio-economic status, or mother’s education. However, non-African-American teachers had lower expectations than did African-American teachers for African-American students, with larger effects for male students and maths teachers.

By conducting this study, researchers hoped to encourage teacher training and professional development to include discussions about expectations and bias, to provide evidence that a more diverse teaching force is needed, and to inform other researchers who look at teacher predictions.

Source: Who believes in me? The effect of student–teacher demographic match on teacher expectations (2016), Economics of Education Review.

More time in class benefits the best

Spending more time at school benefits the best-performing pupils disproportionately, according to a new study.

The researchers used data from the Early Childhood Longitudinal Study – Kindergarten Cohort (ECLS-K). This included more than 20,000 children who entered 1,000 kindergarten (Year 1) programmes in schools across the US in 1998. Children were given maths and reading tests in the autumn and spring. Because there was essentially random variation in when these tests were delivered, there were variations in the amount of teaching time between the two tests. The researchers used this to analyse the progress made, but also the difference in progress among the different percentiles within the class.

They found that, on average, reading scores increase by 1.6 test score standard deviations (SD) during a standard 250 day school year. However, readers in the bottom 10% increased by only 0.9 test score SD, while those in the top 10% increased by 2.1 test score SD. A similar result was found for mathematics. The authors suggest that policy makers, practitioners, and analysts must consider the average and distributional impacts of educational inputs and interventions.

Source: What Differences a Day Can Make: Quantile Regression Estimates of the Distribution of Daily Learning Gains (2015), Institute for the Study of Labor (IZA).