How Does Expansive Framing Promote Transfer? Several Proposed Explanations and a Research Agenda for Investigating Them

Abstract

When contexts are framed expansively, students are positioned as actively contributing to larger conversations that extend across time, places, and people. A set of recent studies provides empirical evidence that the expansive framing of contexts can foster transfer. In this article, we present five potentially complementary explanations for how expansive framing may promote transfer and outline a research agenda for further investigating them. Specifically, we propose that expansive framing may: (a) foster an expectation that students will continue to use what they learn later, which may affect the learning process in ways that can promote transfer; (b) create links between learning and transfer contexts so that prior learning is viewed as relevant during potential transfer contexts; (c) encourage learners to draw on their prior knowledge during learning, which may involve them transferring in additional examples and making generalizations; (d) make learners accountable for intelligently reporting on the specific content they have authored; and (e) promote authorship as a general practice in which students learn that their role is to generate their own solutions to new problems and adapt their existing knowledge in transfer contexts.

ACKNOWLEDGMENTS

This article is based upon research supported by a National Science Foundation CAREER grant to Randi A. Engle (NSF #0844910), a UC Berkeley Chancellor's Fellowship for Graduate Study awarded to Diane P. Lam, and a graduate fellowship from the Research in Cognition and Mathematics Education training grant at UC Berkeley funded by the Institute for Education Sciences (IES #R305B090026) to Sarah E. Nix. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors and do not necessarily reflect the views of any of the funders.

An earlier version of this article was presented in a symposium entitled “Theoretical and Empirical Accounts of Framing in Classroom Interactions” at the Annual Meeting of the American Educational Research Association in April 2011. We appreciate feedback on earlier drafts of this article from Kevin Klein, Adam Mendelson, Kevin Leander, and the current Framing and Transfer research team, which includes Norielle Adricula, Anusha Challa, Jim Clark, Lloyd Goldwasser, Richard Hsu, Christina Lin, Erica Naves, Sarah Perez, Hernan Rosas, Sadaf Sareshwala, Danny Tan, Mai Nhia Vang, Sue Wang, Lynn Yu, and Kathleen Zheng.

Note. Order of authorship is alphabetical as this was a true collaboration, with all four authors contributing equally.

Notes

We want to make it clear that when we say “social context” we are not referring to “problem contexts,” the cover stories in which mathematical and other school-like problems are sometimes expressed (e.g., Goldstone & Wilensky, 2008; Wagner, 2006). Instead, social context is the socially established who, when, where, how, and why of a learning or transfer situation. Given that, in this article, “decontextualization” means somehow removing this surrounding social context so that only the “content” to be learned and (we hope) transferred remains. Part of what would remain after this kind of decontextualization would be any associated problem contexts.

Because processes may or may not occur in particular cases, a dashed line surrounds the box that encloses them.

We note that this explanation differs from Explanation 1, an expectation for transfer, as the expectation for transfer is more of an individual internal one and this involves social expectations. Also an individual expectation for transfer arises from settings being framed as connected and this social expectation for transfer arises from students being framed as authors of particular content that they have learned.

This is the same teacher who was analyzed in Engle et al. (2010), but the teacher's practices have become more expansively framed in several ways from collaborating with the researchers and learning about their findings.

We thank Rob Goldstone for this suggestion.

Evidence of this sort is currently being generated (Meyer, 2012). Preliminary results of exploratory correlational and regression analyses (Engle, Meyer, & Chong, 2012) showed correlations between three different measures of transfer and students’ responsiveness to 10 different instances of expansive framing that spanned both roles and the three main aspects of setting (time, place, and participants). Most correlations were positive and involved students reporting that they had adopted the framing in their actions rather than just agreed with it. Knowing the content to be transferred was correlated with two of the three measures of transfer while many potential predictors of learning like prior grades, test scores, motivational variables, and standard demographic variables did not correlate with transfer. The best regression models all involved expansive framing as predictors and were able to account for between 26% and 60% of the variance in transfer. However, these results are preliminary as we have not yet been able to include in our analyses useful measures of each student's exposure to classical transfer mechanisms like generalizing and comparing examples.

Perhaps a way to get the best of both extremes would be to specify a few key transfer contexts for the knowledge being learned and then specifically open up the possibility of additional contexts with phrases like “and beyond” or “and more.” If done in a believable way, this phrasing may also prompt students to think about additional contexts in which they can use what they are learning.