Recently released Compass scores for Louisiana teachers highlight discrepancies between objective measures of teacher effectiveness and subjective teacher evaluations by principals.
State Education Superintendent John White’s comments on the unevenness of principal reviews from district to district reveal a misunderstanding by White of how validity and reliability work in evaluation.
Assume that the Value Added Measure (VAM) is a valid, reliable way to assess teacher performance, taking into account expected student academic growth, attendance, disabilities, poverty, etc. (I’m not defending VAM here, but White wholeheartedly endorses VAM.)
If VAM is an accurate assessment tool, we must accept that VAM provides a standard against which to compare subjective teacher evaluations by principals. If VAM is significantly higher than principal evaluations, principals are too hard on teachers. If it’s lower, principals are too soft.
We’ll know principals are correctly evaluating teachers when their evaluations coincide with VAM.
White doesn’t seem to grasp this. Both East Baton Rouge Parish School System’s VAM scores and principals’ evaluations indicate 25 percent of teachers are highly effective. Despite agreement between the scores, White concludes that EBRPSS should “adopt an approach that has a high bar for classroom excellence.” He provides Ascension Parish as an example of a high bar.
But VAM indicates 32 percent of Ascension’s teachers are highly effective, while principal evaluations put the number at 6 percent. To a researcher, these results do not mean Ascension principals “set a high bar”; it means they are inaccurately harsh. Zachary and Central, where principal evaluations far exceed VAM scores, lack rigor in their principal evaluations. If accurate, VAM reveals that Ascension was too hot, Zachary and Central were too cold, but EBRPSS was just right.
This is a problem for White. He can’t possibly admit EBRPSS is right about teacher effectiveness.
How can EBRPSS schools be failing if VAM says 62 percent of its teachers are effective or highly effective? How can he justify plans to seize Mayfair Middle and Delmont Elementary because they were “failing” when his Compass system says 25 percent and 42 percent of their teachers were highly effective? He can’t raise VAM standards — that would create a problem for Central and Zachary, two top-ranked districts with VAMs lower than EBRPSS.
This is the problem Compass, in particular VAM, reveals with the school and district evaluation method.
If VAM is an objective, valid method for evaluating teachers, then student characteristics matter. If we know that student characteristics matter in predicting student achievement, why do we account for it in teacher evaluations but not in our school and district performance measures?
Contradictions like this will continue as long as the majority of the public just accepts White’s statements and policies without stopping to consider whether they make sense.
Belinda Davis, Ph.D.
One Community One School District