Teacher-Performance Scores Primed for Release in Virginia
The Old Dominion state is the latest to turn over data on teacher performance as measured by student achievement, this time to a parent who requested it under an open-records law.
Loudoun Now reports that a state judge ruled last month that the state must turn over growth data on Loudoun County, Va., schools and teachers to a parent, Brian Davison, who has sought the information for several years.
A state law in 2013 made "teacher performance indicators and other data" used to evaluate teachers confidential, except persuant to a court order. Davison has published some teacher data from the state in the past, but teachers' names were redacted. The judge ruled that the district didn't show enough evidence as to why teachers' names should be withheld under the state's freedom of information laws.
Davison plans to put the new data up on a Facebook page he's established, though he told the news site that he might shield the names of the teachers that are supposedly the worst performing.
As to the data in question, Virginia uses a "student growth percentile" approach to measuring improvements in student performance. The formula compares the progress of individual students to those who performed similarly on tests in the past. The state does not require districts to use this information in teachers' evaluations. (The Loudoun County district gives teachers and evaluators a lot of flexibility on how to measure student academic progress.)
Aside from the various ethical issues involved in the release of this data, which we've written about before, there's another problem here: Student growth percentiles like Virginia's don't control for factors like student poverty or other demographics that can potentially skew scores.
This isn't merely a pie-in-the-sky worry: Scholars comparing different ways of looking at teacher effectiveness have found that the SGP approach is more affected by classroom composition than other methods of gauging teaching performance. That means the Virginia scores might be reflecting students' demographics rather than how well the teachers are actually teaching. The controversial value-added method, by contrast, explicitly controls for student demographics.