The teacher-evaluation debate follows a well-worn path: Traditional evaluation systems (in which upwards of 99 percent of educators are deemed “effective”) are meaningless, argue reformers. New models that rely heavily on value-added test-score data are unreliable and unfair, counter others. This new NBER working paper from Northwestern’s C. Kirabo Jackson provides the debate new turf on which to tread: Based on data from the 1988 National Educational Longitudinal Study, Jackson channels Paul Tough to argue that students’ “non-cognitive” abilities (adaptability, self-restraint, motivation) help explain their success. Noting this, teachers should be evaluated on them; yet they are rarely considered by current metrics. The report has two parts. First, Jackson shows that the “non-cognitive factor” (which he proxies with variables like absenteeism, suspensions, and grades) is predictive of college enrollment and lifetime earnings—more so, in fact, than cognitive ability. Jackson then evaluates whether teachers can affect this “non-cognitive” factor. Using 2005-10 North Carolina data, he finds that teachers’ impact on student test scores is only weakly associated with their impact on improving youngsters’ non-cognitive abilities. In other words, evaluations that rely exclusively on test scores fail to capture the full breadth of teachers’ contributions to student outcomes. Jackson concludes: Other variables that assess ability to improve students’ non-cognitive skills, variables such as student suspensions and absences, should also be used in teacher assessments. And the debate marches on.
SOURCE: C. Kirabo Jackson, "Non-Cognitive Ability, Test Scores, and Teacher Quality: Evidence from 9th Grade Teachers in North Carolina" (Chicago, IL: Northwestern University, January 2013).