Pay for Percentile

See allHide authors and affiliations

Science  26 Oct 2012:
Vol. 338, Issue 6106, pp. 444
DOI: 10.1126/science.338.6106.444-b

What gets measured gets done. So claim advocates of high-stakes academic testing, arguing that paying teachers on the basis of student performance can improve education. Opponents of “pay for performance” argue that such a system pressures those being measured to game the system, which distorts the process being monitored. High-stakes assessments commonly use similar test topics, items, and formats to maintain consistency of the rating scale over time. This provides opportunity and incentive to “teach to the test,” rather than improving student understanding and achievement. Barlevy and Neal detail an approach to motivate teachers on the basis of student test performance, while limiting the opportunity to teach to the test by using completely new tests each time. With each new test, a student's achievement is reported as an ordinal ranking among a cohort of similarly achieving, similarly tested peers in the school system. A teacher's performance is measured as the sum of her/his students' individual percentile ranks among their respective peer comparison groups. The authors show how “contests” among teachers based on this summed ranking can elicit efficient teacher effort in every classroom.

Am. Econ. Rev. 102, 1805 (2012).

Navigate This Article