In my blog on Monday discussing the Value Added report, I included this quote from one of the authors:
Professor Chetty acknowledged, “Of course there are going to be mistakes — teachers who get fired who do not deserve to get fired.” But he said that using value-added scores would lead to fewer mistakes, not more.
On a more complete reading of the study I found the following information regarding a policy recommendation vis a vis personnel.
“Errors in personnel decisions must be weighed against mean benefits”
The authors argue that if you remove the bottom 5% of the teachers based on VA, and replace them with teachers who are in the median group, student performance and life experience will be significantly better.
Since, however, the accuracy of Value Added metrics increases with more years of data from a teacher, some of that 5% in the study have had only three years of testing in their data, and so they should have more years before we dismiss them.
Using that possibility the researchers argue that the money lost for the students at 28 years old would be close to $400,000. Yes we would be more accurate but at what cost?
“ The marginal gains from obtaining one more year of data are outweighed by the expected cost of having a low VA teacher on the staff even after the first year (Staiger and Rockoff 2010)”
Nota Bene:The authors also acknowledge that simply using VA metrics for personnel decisions could be problematic and that principals’ observation and evaluations would provide good additional information in those decisions.
