Value-added formulas that aim to assess teacher effectiveness produce different lists of “best” and “worst” teachers depending on whether the formula controls for student background, according to a new study published in the journal of the American Statistical Association.
The study found that VAM models that do not control for student background disproportionately rank teachers in low-income schools in the bottom quartile and disproportionately rank teachers in well-off schools in the top quartile. By contrast, VAM models that incorporate student demographics — including eligibility for subsidized lunches, race, gender, special education status and English-language learner status — produce rankings in which teachers from advantaged and disadvantaged schools are evenly represented throughout each quartile.
The study found only “modest benefits” from using multiple years of student test performance in a VAM formula, as opposed to a single year. Using a single year did not give any particular type of school or teacher an advantage.
The researchers, all economists at the University of Missouri, conducted the study with data from middle-school math and language arts test scores.
No comments:
Post a Comment