KNOWLEDGE BRIEF 12
by Dan Goldhaber
Teacher training programs are increasingly being held under the microscope. Policymakers are beginning to adopt the use of student growth measuresA student growth measure is a measure of the amount of academic progress a student makes between two or more points in time. to measure the performance of teacher preparation programs. In this brief, Goldhaber discusses what research can tell us about the usefulness of value-added for assessing the quality of teacher preparation programs.
KNOWLEDGE BRIEF 12
KNOWLEDGE BRIEF 11
by Douglas N. Harris
States and districts across the country are using multiple measures to make personnel decisions about teachers based on a weighted averageA weighted average is a way of computing an average out of a set of numbers in which some of the numbers in the set carry more importance (weight) than others. of the separate measures. This method has strengths and weaknesses. Most discussions of measures of teacher performance measures focus on validityThe degree to which something measures what it claims to measure. and reliabilityReliability refers to the degree to which the measure is consistent when repeated. ; but fairness, simplicity, and cost should also be considered.
KNOWLEDGE BRIEF 10
by Stephen Raudenbush
Statistically, comparisons of teachers across different schools make less sense in districts with high levels of between-school segregation.
Much of the research on the validityThe degree to which something measures what it claims to measure. of value-added for teacher evaluation has been based on studies of its use to rank teachers within the same school. However, most districts are using value-added to rank teachers across the district, and in districts that reflect neighborhood residential segregation, value-added rankings will compare teachers who teach very different types of students. To try to correct this by statistically controlling for school context variables will produce biasThe systematic difference between an estimated or expected value and its true value. so long as the effectiveness of the teacher is associated with attributes associated with the effectiveness of the school.
KNOWLEDGE BRIEF 9
by Susanna Loeb and Jason A. Grissom
As policymakers begin to consider how to use test scores to assess principals, they should be guided by an understanding of how principals influence student outcomes. Value added measures provide information about how schools are doing, but they may not be convincing measures of the causal effect of the principal of student learning.read more »
KNOWLEDGE BRIEF 8
by Daniel McCaffrey
A teacher’s value-added can depend on the test used to assess his or her students’ achievement. Questions about the effect of a test on a teacher’s value-added are particularly salient today because most states will soon be adopting new tests aligned with the Common Core State Standards. Research and past experience suggests that teachers should be prepared for greater year-to-year variability in their value-added ranking for a few years after the change when districts use both old and new tests for value-added calculations.read more »
KNOWLEDGE BRIEF 7
by Douglas N. Harris
Value-added methodology is being applied to the evaluation of teachers in tested grades and subjects, but the vast majority of the research on value-added measures focuses on elementary schools only. Secondary grades differ from elementary grades in ways that are meaningful for the validityThe degree to which something measures what it claims to measure. and reliabilityReliability refers to the degree to which the measure is consistent when repeated. of value-added measures for secondary teachers. Middle and high school teachers have more students, which increase reliability of value-added scores, but this advantage is offset by tracking, which reduces reliability at those grade levels.read more »
What Do We Know About the Tradeoffs Associated with Teacher Misclassification in High Stakes Personnel Decisions?
Evaluators have to rely on inherently imperfect measures to rate teachers. As a result, evaluating teachers to group them into performance categories will inevitably lead to errors. Errors result in “false positivesA false positive, or false identification, is the classification of a teacher into a group that he or she does not actually belong. For example, classifying a teacher as "not effective", as measured by value-added scores, who is actually "highly effective" is a false positive. This phenomenon is due to statistical error in the modelA model is a statistical equation describing the variation in the outcome of interest when one or more other variables change. . ” and “false negativeA false negative, or false non-identification, is the failure to classify a teacher into a group that he or she does actually belong. For example, failing to identify a teacher as "not effective", as measured by value-added scores who is really ineffective is a false negative. This phenomenon is due to statistical error in the model. ” classifications, which have important implications for students and teachers.read more »
KNOWLEDGE BRIEF 5
by Douglas N. Harris
In the recent drive to revamp teacher evaluation and accountability, teacher value-added measures have unquestionably played the starring role. But the star of the show is not always the best actor, nor can the star succeed without a strong supporting cast. In assessing teacher performance, observations of classroom practice, portfolios of teachers’ work, student learning objectives, and surveys of students are all possible additions. In this paper, I will explain how these various measures stack up on two essential criteria: validityThe degree to which something measures what it claims to measure. and reliabilityReliability refers to the degree to which the measure is consistent when repeated. .read more »
Given the modeling and vendor options at their disposal, school districts and states likely have a number of pressing questions about which modelA model is a statistical equation describing the variation in the outcome of interest when one or more other variables change. is “right” for their specific situation. This entry explores the amount of difference choice of model makes. How would the same teacher rank under different modeling approaches? And, in particular, in what effectiveness category would the same teacher fall under different modeling approaches?read more »
Value-added measures are being used to assess teacher effectiveness, but how can we make sense of the inconsistency in value-added measures for the same teacher across time, subject and student population? Some of the inconsistency in a teacher’s value-added measures is driven by true differences in that same teacher’s performance. That is, a teacher may simply perform better in one year than he or she does in another year. Another part of the difference comes from the inaccuracy of the value-added measure. Understanding the variationThe extent to which a set of numbers is spread out. in value-added measures for the same teacher across time, subjects and students can help education leaders make the best use of the information that value-added measures provide on teacher performance to inform the decisions they make.read more »