Student growth modeling and teacher value-added modeling are complex due to non-random assignment, measurement error, multidimensionality, and covariates¶

Claim · All claims
Evidence · 1 study · theoretical r? · q1
Currency · newest study 2015 · not yet checked against its sources · not reviewed by a person · overdue for a recheck since 2018: tool and model accuracy changes quickly

Subclaims¶

q1 i? The complexity of student growth modeling and teacher value-added modeling stems from non-random assignment of students, measurement error in achievement scores, multidimensionality of the measured construct across grades, and the inclusion of covariates. → The effective use of student and school descriptive indicators of learning progress: From the conditional growth index to the learning productivity measurement system 2015

Evidence¶

The effective use of student and school descriptive indicators of learning progress: From the conditional growth index to the learning productivity measurement system 2015¶

The effective use of student and school descriptive indicators of learning progress: From the conditional growth index to the learning productivity measurement system. (2015). Value Added Modeling and Growth Modeling with Particular Application to Teacher and School Effectiveness, 237–269. https://www.nwea.org/research/publication/the-effective-use-of-student-and-school-descriptive-indicators-of-learning-progress-from-the-conditional-growth-index-to-the-learning-productivity-measurement-system/ doi:10.1108/978-1-62396-776-5

q1 · i? · theoretical · r?

The article's abstract states this as a framing claim about why "Student growth modeling and teacher value-added modeling are complex", enumerating four sources: non-random assignment, measurement error, multidimensionality across grades, and covariates. No data or analysis supports it in the available text.

"The complexity stems, in part, from issues due to non-random assignment of students into classes and schools, measurement error in students’ achievement scores that are utilized to evaluate the added value of teachers, multidimensionality of the measured construct across multiple grades, and the inclusion of covariates."

Discussion¶

Related Claims¶

• Multidimensional IRT scores that account for latent variable covariances over time produce better recovery of growth parameters in longitudinal growth models than sum scores and other IRT approaches — related
• Defining and quantifying reliability and measurement error is particularly challenging in epistemic games for three principal reasons — related
• The Rothstein test can reject VAMs even when students are randomly assigned conditional on the covariates in the model — related
• The Memphis value-added model estimates school effects on student performance while adjusting for mobility, demographics, and measurement error — a narrower finding that bears on this claim
• Value-added error rates apply to measuring both teacher and school performance in the upper elementary grades using student test score gain data — related
• A low alpha for an error-prone covariate can attenuate, inflate, or artificially produce a spurious effect (phantom effect) in the coefficient of a perfectly reliable predictor — related