How much formalization of assessment methods is useful when implementing formative assessment in second grade mathematics classrooms?

Student achievement has been found to benefit from formative assessment (FA), but only few studies have directly compared different FA approaches. In the present study, two differently structured FA approaches were compared: While teachers in the first experimental group (curriculum-embedded assessment (CE), 19 classes, N = 431students) used written task sets for assessment, teachers in the second experimental group (planned-for-interaction assessment (PI), 22 classes, N = 492 students) were trained in conducting focused conversations for assessment with their students throughout the school year. Results suggest a slight advantage of CE in terms of math achievement, whereas PI was associated with significantly higher scores on the variable need for cognition. Students’ academic self-concept did not differ between groups, and results were not affected by students’ achievement level or teacher characteristics. Consequently, it seems promising to combine the highly structured CE approach with the more flexible PI approach.