Teacher Questioning: Opportunities for Retrieval Practice in Middle School Math Classrooms

• In Event: Investigating Cognition and Motivation at the Classroom Level

Abstract

Researchers now recognize the beneficial effects of retrieval practice as a study strategy. Activities that involve retrieving information from memory such as answering test questions and recalling a text are more effective at improving learning than restudying, concept mapping, and other study techniques (Karpicke & Blunt, 2011). These benefits occur on both verbatim and transfer tests (Butler, 2010), when students do and do not receive feedback (Butler & Roediger, 2008), and in laboratory and classroom studies (Roediger, Agarwal, McDaniel, & McDermott, 2011).

These studies suggest that retrieval practice opportunities provided by teachers during lectures and classroom discussions should also improve learning, but this idea has not been empirically tested (Karpicke, 2012). Preliminary evidence with 1st and 2nd grade classrooms suggests that teachers who included both retrieval practice and metacognitive talk in their teaching had students who were later more strategic in solving problems and had larger increases in math fluency and calculation (Grammer, Coffman, & Ornstein, 2013; Grammer, Coffman, Sidney, & Ornstein, in press). However, the sample size for the math study was small, the math topics were fairly simple, and the school district was suburban and high-performing.

We examined teacher questioning in urban middle schools where the math topics are more difficult, the teachers less experienced, and the students typically less prepared for challenging tasks (Darling-Hammond, 2010). We asked two main questions 1) Do teachers naturally use retrieval practice when talking to their students? and 2) Is the amount of retrieval practice related to student learning? To answer these questions we took advantage of the Measures of Effective Teaching longitudinal database which features classroom videos and measures of student learning. We restricted our sample to 40 6th – 8th grade classrooms.

In preliminary analyses of 14 classrooms we have seen wide variation among teachers in their use of questions with 15% to 40% of teacher utterances being math relevant questions (M = 23%). All teachers provided opportunities for retrieval practice, but they varied substantially in the frequency of such questions (23% - 68% of math-relevant questions provided an opportunity for retrieval, M = 44%). There appears to be a relation between the proportion of teacher utterances that were opportunities for retrieval practice and teachers’ value-added scores on a conceptually-based test of math achievement, r(12) = .30, p= .31. However, visual inspection of the graph shows two outliers where teachers had higher than predicted value-added scores given the proportion of retrieval opportunities. Excluding these two classrooms, the relation was much stronger, r(10)= .84, p < .01.

Our results suggest that urban middle-school math teachers often provide opportunities for retrieval practice in their classroom discourse and that the frequency of those questions is related to student learning. This relation was found not with verbatim repetition of questions, as is typical in studies of retrieval practice, but between teachers’ natural discourse and students’ conceptual knowledge of mathematics. The results provide further evidence that the benefits of retrieval practice are broad and long-lasting.

Author

• Lisa Fazio, Vanderbilt University