Learning in Teacher Partnerships: Experimental Evidence From the Instructional Partnership Initiative in Tennessee

• In Event: District and State-Level Approaches to Teacher Improvement

Abstract

We study how teachers learn on-the-job, especially how they learn from coworkers. We examine data from a field experiment in Tennessee. Treatment schools were invited to participate in the Instructional Partnership Initiative (IPI), which is designed to leverage the feedback teachers receive during formal evaluation into lasting improvements through collaborative partnerships. Pairs are matched on specific skills measured in prior evaluations. Each pair includes a target teacher who scores low in one or more specific skills, and a partner who scores high in the same skills. We measure whether teacher performance improves when two teachers working at the same school are paired and asked to work together on improving their skills.

Theoretical framework

One hypothesis for why existing development programs often fail to improve teaching is that the typical program is a one-size-fits-all approach, not tailored to the needs of individual teachers (Garet et al., 2011). Studies that do find positive effects typically include sustained work and focus on teachers’ individual, current instructional challenges (Desimone et al., 2002). The IPI combines a focus on individual teacher needs with individualized, local assistance from a colleague.

Methods and data sources

In 2015, the Tennessee Department of Education began a statewide randomized experiment of the IPI with over 1000 schools statewide. We measure the treatment effects of IPI on multiple outcomes: student test scores, teacher classroom observations, statewide surveys of administrators and teachers. We also examine differences in schools’ participation in IPI. The data are primarily administrative records and existing survey data, with specific survey questions about the IPI program.

Results

In a pilot study experiment, we document improvements in the performance of teachers participating in IPI (Authors, 2016). Student test scores improved by 0.12 standard deviations in target teachers’ classrooms. These improvements persisted, and perhaps grew, in the year after treatment. Empirical tests suggest the improvements are likely the result of teachers learning new skills from their partner. While these pilot results are encouraging, measuring program impacts statewide across the diversity of the state’s schools is important to further our understanding of whether and how teacher partnerships influence learning and performance. This presentation will report results of the statewide field experiment, which began in 2015. Spring 2017 tests are the first opportunity to measure IPI’s effect on student achievement statewide because spring 2016 tests were largely cancelled in Tennessee.

Scholarly significance

Learning new skills on-the-job from coworkers is an intuitive method of human capital development in many occupations, but has received little empirical attention. This is the first study, of which we are aware, to examine such learning using experimental variation in treatment and direct measures of job performance. For schools specifically, the results contrast a largely discouraging lack of performance improvements generated by formal professional development training for teachers.

Authors

• Eric Taylor, Harvard University

• Ellen B. Goldring, Vanderbilt University

• Jason A. Grissom, Vanderbilt University

• Mary Elizabeth Laski, Brown University

• John Papay, Brown University

• Susan Kemper Patrick, Vanderbilt University

• John H. Tyler, Brown University