Impact of Teacher Observation Scores, Value-Added Scores, and School Environment on Retention

• In Event: 68.029 - Key Teacher Evaluation/Accountability System Components: Using Diverse Methods to Inform Practice and Policy

Abstract

Purpose

While wide-scale federal and state legislation has incentivized high-stakes testing systems meant to hold teachers accountable for effectiveness in the classroom, based on the value they add to student test scores, there has been less attention paid to legislative/system impacts on simultaneous teacher retention rates. In this paper researchers examine the relationships among teacher observation, student achievement as measured through value added scores, and working condition data from one large-scale system.

Perspective

Teacher turnover inherently diverts both monetary and human capital resources away from student learning. Teacher turnover further disrupts the development of professional communities, contributes to organizational instability and uncertainty, and interrupts implementation of educational reforms (Ingersoll, 2001). Accordingly, rates of teacher turnover are also higher in schools that serve high-poverty and minority students, and subsequently in schools where teacher quality is more paramount (Ingersoll, 2002; Ronfeldt, Lankford, Loeb, & Wyckoff, 2013).

Methods

Researchers examined data from 2,109 teachers, including all teachers’ classroom observation scores (i.e., the TAP instructional rubric, with observations across multiple domains four times per year; NIET, 2011), value-added scores (i.e., the Student Growth Percentile, SGP, package), and retention information from year to year. Self-reported working conditions data (i.e., using the North Carolina Teacher Working Conditions Survey, The Center for Teaching Quality, n.d.) were also available for 1,509 (71.5%) of the total 2,109 teachers. Hierarchical regression was utilized to examine the impact of working conditions and years of teaching experience on teacher observation scores and median SGP. Logistic regression using blocks of data (teacher observations, classroom median SGPs, and working condition scales) examined patterns of teacher retention.

Results

When examining what variables were associated with median SGP, taking into account teacher working conditions and years of experience, teacher observation scores was the only variable significantly predictive of median SGP (B=.40, p < .001). When the model was examined using teacher observation scores as the outcome variable, perceptions of school facilities, years of experience, and median SGP were all significant predictors, even in the full model (R2 = .23, p < .001).

Extending this finding, logistic regression examined whether these same variables impact teacher retention. Researchers found that higher teacher observation scores significantly increased the likelihood of a teacher being retained (Exp (B) 2.22; p<.01) above and beyond median SGP in the model. When working conditions were included in the model, taking into account teacher observation scores and median SGP, only time remained significantly associated with retention (Exp (B) 1.51; p<.01).

Significance

This study highlights the complex nature of teacher evaluation systems and the impact on retention. While this study shows that effective teachers, measured by teacher observation, are staying in the classroom at higher levels, no impact of Median SGPs, the value-added model used in this study, was apparent in the model. Teacher perception of working environment, in particular facilities available, plays a role in teachers’ decisions to remain in their school. Thus, it is crucial to use multiple measures of teacher effectiveness along with perceptions of working conditions to predict retention of effective teachers.

Author

• Wendy Miedel Barnard, Arizona State University