How Teachers Work Together and How It Matters: The Relationships Among Teacher Social Capital, Teaching Practices, and Student Achievement

• In Event: Promising Scholarship in Education: Dissertation Fellows and Their Research

Abstract

Purpose of study
Compelling evidence suggests we ought to pay attention not just to who works in our schools, but to how those people interact. Effective schools are characterized by high levels of social capital – a trusting environment in which teachers collaborate and rely on one another for instructional advice (Bryk & Schneider, 1996; Leana & Pil, 2006). However, research has not yet detailed the mechanisms through which social capital benefits students. In addition, much available research is limited to single-level analyses of manifest variables, which may result in biased estimates (Lüdtke, Marsh, Robitzsch, & Trautwein, 2011). This study takes advantage of the rich data available in the Measures of Effective Teaching (MET) project and applies multilevel structural equation modeling to account for the nested data structure and the measurement error in the observed indicators.

Study design and methods
This study seeks to contribute to our understanding of the relationships among teacher social capital, teaching practice, and student achievement (see Figure 1). To do so, I investigate the following research questions:

• How does teacher social capital relate to teaching practices and student achievement? How do these relationships change based on the use of different measures of teaching practice (observer ratings versus student survey ratings)?
• How do the relationships among these constructs differ depending on the choice of model (a multilevel structural equation model or models that ignore either clustering or measurement error)?

In terms of measures, the preliminary results reported here focus on social capital. Teacher social capital is conceptualized as a construct comprising four dimensions, measured by Teacher Working Conditions Survey (TWCS) items, and operationalized as follows:

1) Networks (teacher-teacher, -administrator, and -parent). Strength is defined operationally by the frequency of interactions and the obligations and expectations shared by the people in relationship. TWCS measures the time available for teachers to collaborate and the perceptions of communication between parents and teachers;
2) Norms of a school (trust, respect, and caring). Norms are operationalized through teachers’ perceptions, the extent to which they feel trusted (and respected and valued), and the extent to which they trust other teachers and administrators to be competent and to take positive action on behalf of the students;
3) Expertise/resources/human capital available to teachers through their networks. Operationally, this is measured by teachers’ reports of the extent to which they receive meaningful feedback and have access to support from colleagues, administration, support personnel, and professional development opportunities; and
4) Depth of interactions taking place within a teacher’s network. Depth of interactions is measured through teachers’ reports of the extent to which they collaborate on meaningful instructional activities such as assessment, lesson planning, and alignment.

Teaching practices will be measured through the Tripod survey and the Framework For Teaching and conceptualized as comprising the dimensions of teaching support and teaching press. Student achievement will be measured through state math tests and the Balanced Assessment of Mathematics.

Multilevel structural equation modeling accounts for the nested structure of the data and the latent nature of the constructs of interest (see Figure 2). Consequently, these models control for measurement error coming from the sampling of items at level one (students or raters) and level two (classrooms/teachers). These models also account for the fact that students’ responses on surveys are clustered within classrooms and observers’ ratings on rubrics are clustered within teachers. The models will be compared to alternative and often employed approaches that ignore either clustering or measurement error.

Preliminary findings and policy implications
The conceptual model of this study presumes interrelationships among social capital, teaching quality, and student achievement. Patterns of correlations among items indicate the four dimensions of teacher social capital can be measured adequately by the TWCS. The poster will include results of additional analyses currently under way investigating confirmatory factor analysis models to confirm these indications.

Ultimately, this study is expected to yield insight into illustrating the mechanisms through which social capital may impact teaching practices and students’ achievement. In addition, this study will shed light on how our inferences about the relationships among teacher social capital, teaching practices, and student achievement are affected by the choice of measures and models.

Author

• Kevin Schaaf, University of California - Los Angeles