Theorizing and Measuring Social Supports for Teacher Learning

• In Event: Measuring and Supporting the Improvement of Common Core State Standards–Aligned Mathematics Teaching at Scale

Abstract

Objective
The objective of this paper is to provide early evidence of the viability of a theory of social supports for the improvement of mathematics teaching, including how it relates to instruction as characterized by the quadrant-theory introduced above. Practically, Tennessee has invested heavily in instructional coaches; the extent to which teachers’ access to coaches and other forms of expertise and supportive collegial interactions predicts instructional quality will constitute an essential component of our suite of measures, thereby aiding leaders’ decision making regarding how to deploy and distribute coaches in upcoming years.

Theoretical Framework
There is growing consensus in the research literature that the informal interactions that occur in teachers’ school-based professional communities are critical to teacher learning and development (Bryk et al., 2010; McLaughlin & Talbert, 2006). This research suggests that teachers’ social interactions provide access to instructional expertise (Frank, et al., 2004; Penuel et al., 2009), enable them to learn from one another (Gersten et al., 2000; Klingner et al., 1999), and provide normative pressure to engage in new practices (McLaughlin & Mitra, 1999). Drawing on a network perspective (Borgatti & Ofem, 2010; Daly, 2010), recent research has sought to more precisely examine the role of professional interactions in instructional improvement, pointing to key dimensions of teachers’ networks including the strength of ties, access to expertise, and the depth of interaction (Frank et al., 2004; Obstfeld, 2005; Penuel et al, 2009; Reagans & McEvily, 2003).

Methods/Data sources
Guided by network theory, we aimed to measure four key dimensions of teachers’ support networks: tie span (degree to which networks include administrators, coaches, and grade level peers); tie strength (frequency of interaction); expertise related to mathematics teaching and learning; and depth (extent to which interactions take up substantive issues related to teaching and student learning). We designed a survey to measure these constructs supported by more intensive forms of data, such as, interviews, interaction logs, and videotapes of teachers’ collegial interactions. Drawing on the full corpus of our data, we identified profiles of teachers with qualitatively different patterns of collegial interaction and related these profiles to different patterns of teaching as described by the quadrant theory.

Results
Our exploratory analysis identified the following features of teachers social networks to be associated with high-concept/high-struggle teaching (Q1): access to a coach with relatively high expertise; higher proportions of high-depth interaction with grade level colleagues; and more frequent interaction with a relatively expert administrator. In contrast, teachers that engage in low-concept/low-struggle teaching (Q4) were more likely to have less frequent and less in-depth interactions with teaching colleagues, less frequent interaction with coaches, and less frequent interaction with more expert administrators.

Scholarly Significance
Our analyses contribute to the extant literature by relating features of teachers’ social networks to different instructional profiles. In addition, the findings have practical significance, providing suggestive evidence that we can measure important features of social supports with survey measures deployable at scale, which provide actionable information to state leaders as they monitor and direct resources to further instructional improvement efforts.

Authors

• Jennifer L. Russell, University of Pittsburgh

• Mary Kay Stein, University of Pittsburgh

• Richard James Correnti, University of Pittsburgh

• Stacy Song Kehoe, University of Pittsburgh

• Debra W. Moore, University of Pittsburgh