Wednesday, August 31, 2016
New Jersey’s Principal Evaluation System
A new study finds both strengths and limitations in the measures used for New Jersey’s new principal evaluation system. The study, released by the Institute of Education Sciences today (Aug. 30), was conducted by Regional Educational Laboratory Mid-Atlantic and focused on the statewide principal evaluation system that was fully implemented by the state of New Jersey during the 2013-14 school year and piloted in 14 districts during the 2012-13 school year.
New Jersey principals were rated on four components: two measures of principal practice and two measures of student achievement. The student achievement measures included goals for school achievement and teachers’ student growth objective averages. Principals of schools with grades 4–8 were also evaluated on school median student growth percentiles, a measure of student achievement growth based on state assessments in math and English language arts.
The study examines four statistical properties of the measures: the variation in ratings across principals; year-to-year stability; the associations between component ratings and student characteristics; and the associations among component ratings. Analyses of statewide and pilot data yield several key findings:
• Nearly all principals received effective or highly effective overall ratings in 2013-14. However, fewer principals evaluated on school median student growth percentiles received highly effective overall ratings;
• Principal practice instrument ratings and school median student growth percentiles were moderately stable across years, providing evidence of reliability; and
• Several of the principal evaluation measures, including the overall rating, were slightly lower at schools with more disadvantaged students.
The findings indicate that additional guidance or alternate measures may be needed to better differentiate principals' performance and to more consistently rate principals who are and are not evaluated on school median student growth percentiles. Further research might investigate the causes of the relationships between principal ratings and the percentage of economically disadvantaged students.