Using test-score gains to evaluate teachers: The Gates Foundation hasn't done its homework.

The Gates Foundation hasn't done its homework

Sent to the Seattle Times, June 9, 2013

The Gates Foundation is now willing to "put less emphasis" on test-score metrics in teacher evaluation ("Gates Foundation looking to make nice with teachers," June 8).

The Foundation hasn't done its homework. The research clearly shows that test-score gains don't belong in teacher evaluation at all:  They are unreliable (unstable), giving different results for the same teacher from year to year and different results when different tests are used. They don't work, period.

Stephen Krashen

Professor Emeritus

University of Southern California

Sources:

Different tests produce different ratings: Papay, J. 2010. Different tests, different answers: The stability of teacher value-added estimates across outcome measures. American Educational Research Journal 47,2.

Unreliable: Sass, T. 2008. The stability of value-added measures of teacher quality and implications for teacher compensation policy.  Washington DC: CALDER. (National Center for Analysis of Longitudinal Data in Educational Research.)  Kane, T. and Staiger, D. 2009. Estimating Teacher Impacts on Student Achievement: An Experimental Evaluation. NBER Working Paper No. 14607 http://www.nber.org/papers/w14607;

Original article:

http://seattletimes.com/html/education/2021149398_gatesfoundationteachersxml.html