Few things can start up an argument faster than the role of student standardized tests. They are blamed for narrowing the curriculum and giving the public a false perception of the state of public education. Some think they should be used to evaluate teachers and establish pay scales. Others decry their use even as a measure to evaluate student performance – unless the scores rise.
There are a lot of very smart people out there trying to bring these sides together to craft some sort of system or compromise that would allow student scores to be factored into evaluations of teachers and schools, without losing sight of the many variables that enter into student outcomes. In some locations, they have even succeeded.
I think they are wasting their time.
The debate over performance pay is a good example. Statements from the National Education Association consistently mention opposition to the practice “based on student test scores.” Proponents then figure if they could come up with a system not solely based on student test scores, it might pass muster. They are unaware that NEA is philosophically opposed to performance pay, however it is determined.
Even if we leave the issue of pay aside, these are differences of belief, not of methods. It is official NEA policy that student test scores don’t reveal anything about the quality of teaching or schools. Poor test results indicate a change of tactics may be in order, not that the school and teachers are deficient. NEA believes “that indicators of student learning are most appropriately used in formative assessments focused on helping teachers to improve their practice. Measures of teacher effectiveness based on standardized test data should not be used for summative evaluation of teachers or other education professionals.”
Do “value-added” assessments manage to sidestep this problem? Not according to NEA’s Committee on Professional Standards and Practice:
It should be noted that so-called “value-added” models that claim to measure teachers by student test data do not promote collaboration and can measure teacher effects only when teaching remains a private and isolated process. These measures of isolated practice are neither valid nor reliable.
Under these circumstances, we will continue to see what we have already seen. In most places, student scores will never formally become a part of teacher evaluations. In some places, individuals on opposite sides will craft something they can both live with. But we can rule out a grand consensus.