I found that parents of public school children residing in states with more highly developed assessment systems expressed more negative attitudes about their children’s schools and about government in general than public-school parents in states with less extensive testing policies. I also found that parents in states with more stringent policies were somewhat less likely to participate in their children’s education.
Walk us through this measure of “assessments systems.” What are they, and why would states with highly developed systems also be the states where parents are less involved in their child’s school?
My ultimate objective was to gauge the impact of the “standards-based reform” agenda on parents. Standards-based reform is the idea that state governments can improve schools by adopting centralized standards, testing and accountability policies that measure student performance and reward or penalize schools based on achievement. Assessment policies are central to this paradigm, providing “objective” measures of student progress toward the standards and data-driven grounds for rewarding or punishing schools.
Crucially, in the standards-based reform model, the question is not simply whether states test students, but whether they adopt a suite of “best practices” believed to make the system work properly. My measure of state assessments was a count of the number of policies adopted by the state that related to the standards-based model. The index measured (1) what types of tests states used; (2) the grade levels at which they were employed; (3) whether the assessments were aligned to state standards; (4) whether they were “vertically equated” so that test scores for each grade were comparable; and (5) whether the state provided sample exams and/or test items to educators to facilitate alignment of curriculum and instruction.
While these policies are arguably well-meaning, they may be off-putting to parents. Standards-based reform agenda has been driven primarily by educational interest groups and technocrats, with limited input from public-school parents. So parents may feel disempowered by these reforms.
Second, high-stakes assessment regimes can lead teachers and administrators to adopt undesirable behaviors — teaching to the test, cutting time from non-tested subjects, and, in extreme circumstances, cheating — that create the appearance of high performance. By eroding the quality of education, such activities may frustrate parents, weakening their confidence in government and suppressing their participation.
But how do you know that assessment policies are really a factor? How can you rule out other explanations?
When working with observational data, there is always some uncertainty. I cannot completely eliminate the possibility that some other factor actually explains the effect I ascribe to state assessments.
However, I’ve done my best to alleviate this concern. I account for a large number of individual and state factors that could explain the outcome, and the effect of state assessments persists.
As a further safeguard, I used statistical matching. Roughly speaking, matching prunes observations from the data so that remaining observations — in my study, public-school parents — are basically identical except for whether they live in states with stronger or weaker state assessment policies. Matching allows the researcher to come closer to the ideal of a randomized experiment.
However, matching only adjusts for the factors I have measured, leaving the possibility that factors I haven’t measured may affect whether parents live in states with strong or weak assessment policies and thereby confound my conclusions. As a final check, I examined how sensitive my findings were to the possibility of unmeasured factors. Most of my results were not sensitive to such factors.
There is a lot of debate right now about the Common Core standards. Do your findings speak to this at all?
My analysis predates the implementation of Common Core, so some caution in applying my conclusions is warranted. However, the main takeaway is that states must be aware of the risks of implementing centralized, highly technical assessment policies. These policies can have unintended and undesirable effects on parents’ attitudes about education and government, as well as their participation in their children’s schooling.
My findings may therefore provide some insight on the parental revolt against the Common Core. Basically, parents are feeling disempowered about the increasing centralization and standardization of education, and they are frustrated and angry about it. Over the long run, this backlash could make these reforms unstable from a political standpoint. Since there is some evidence that these policies can induce improvements in average student achievement, parental backlash is something that states need to worry about.
What is the lesson here? Should states change their assessment policies?
There are two broader lessons. First, states need to do a much better job explaining what the assessments are, how they work, why they are important, and how they really benefit children. Standards-based reform policies have been with us since the 1990s, but parents are still baffled by them.
Second, states need to do more to bring parents into the policymaking process. As long as these policies are perceived as being done to (rather than with and for) children and their parents, the parental response will be negative. States need to make parents active participants in the formulation of policies that profoundly influence their children’s lives.
For more, see the article and also Rhodes’s book, An Education in Politics: The Origin and Development of No Child Left Behind.