The role of teacher beliefs and decision practices in the equitable access to educational opportunities in mathematics

Despite gains in academic achievement and attainment among racial and ethnic groups over the last 40 years, the relative performance of students from differing backgrounds is far from equal. The causes for the achievement gap are as much structural as ideological. The gap is rooted in systems of economic and social privilege that have created belief systems and practices resulting in differential treatment of students. National programs aimed at eliminating academic difference in achievement based upon race and other factors have had only limited success (IASA, 1994). Because academic success has been increasingly gauged by standardized tests and assessments, instruments sensitive to social stratification, such tests may have reinforced beliefs of school personnel; new policies, however, have emphasized the use of different indicators of student performance. The study addressed how these different indicators related to teacher beliefs and decisions. The study, conducted in a web environment, investigated the effects of educator background, causal attributions, and student ethnicity (displayed in photographs) on placement decisions. The order of causal attribution and placement decision was counterbalanced. Teachers and administrators were presented with a diverse set of 32 students' photographs paired randomly with data profiles containing four types of indicators (norm-referenced test scores, performance categories, grade point average and homework completion information). Based on the given information, educators were asked to recommend students for a challenging mathematics course. The study found that attribution patterns were similar across indicators. Surprising, and counter to expectations, educators were found to make comparable and strong Effort causal attributions to success for students in all ethnic groups. Educators' Effort attributions were significantly lower for the norm-referenced test indicator. Accuracy of nominations was affected by the variance in attributions. The findings did not support the differential use of information for students of different backgrounds. The study was limited to high-performing student profiles and should be replicated on a wider range of student achievement. If substantiated by future research, the study may indicate that educators are valuing norm-referenced tests less and see them less subject to improvement by student effort.