When the rubber meets the road: putting research-based methods to test in urban classrooms

We created a hypothetically "optimal" instructional scenario in which a knowledgeable researcher, under the guidance of an experienced classroom teacher, carried out a set of research-based science instruction in a low-SES urban school. The training group's performance was assessed by standardized test items and compared with that of a high-SES no training comparison group. The results demonstrate that instructional methods based on experimental psychological research have great potential for addressing the achievement gap problem. However, the analyses also reveal a significant discrepancy between low-SES students' performance on standardized test items and on alternative assessments with lower reading and writing demands. We discuss our methodological choices in making basic research more relevant to real world issues. We highlight the critical challenges of test validity, research relevancy, and reform feasibility for researchers and policymakers.