Choosing the Right Measure for Student Test Score Growth

February 24, 2014

A "two-step" model is the best way to measure student test-score growth, say Mark Ehlert, Cory Koedel, Eric Parsons and Michael Podgursky, economics professors at the University of Missouri, Columbia.

There are three ways to measure test score growth as a way to evaluate teacher and school performance: aggregated student growth percentiles, a one-step value-added model, and a two-step value-added model.

Aggregated student growth percentiles (SGPs) evaluate a student's growth from one year to the next, compared to the growth of his peers.

  • These models do not control for differences in student backgrounds or learning environments.
  • Instead, an SGP simply takes a student's testing score in a current year and compares it with the performance of students who scored similarly to that student in the prior year.
  • An aggregated SGP takes these figures for individual students and combines them for an entire district or school or teacher.

The most common approach to measuring growth is the one-step value-added model, which does control for student background and can sometimes be useful in determining whether schools and teachers are impacting student achievement. The problem with this approach, however, is that it requires assumptions that, if incorrect, can yield erroneous results.

Two-step value-added models first look at the relationship between achievement and characteristics of that student and the school. Then, the model adjusts for differences in school and student characteristics before comparing growth measures across the board. The authors prefer this approach because they say that it encourages optimal effort by schools and teachers. And by "leveling the playing field" in assessing student performance, schools can see how well they are doing compared to similar institutions. Such an evaluation method can also strengthen the ability of high-poverty schools to recruit and retain good teachers, who would otherwise be hesitant to teach in these high-poverty areas.

The study also suggests continuing to report test-score levels along with the information from this two-step approach in order to ensure that policymakers still see the differences in achievement between schools, still providing an absolute measurement of which schools have low test scores.

Source: Mark Ehlert, Cory Koedel, Eric Parsons and Michael Podgursky, "Choosing the Right Growth Measure," Education Next, Spring 2014.

 

Browse more articles on Education Issues