Commentary

Does Assessment Make Colleges Better? Who Knows?

Michael Morgenstern for The Chronicle

August 14, 2015

Last year the younger of my two sons went off to college. As we went through the search process, we looked at university and department websites, checked faculty research interests, looked for evidence of faculty involving students in their research, flinched at the prices, marveled at the climbing walls, and considered quality of the food on campus. Basically we did all the things a typical middle-class family would do in a college search, along with a few insider concerns like looking at faculty publications and grants and checking that the university libraries had at least one of my books. In retrospect one question that never crossed my mind was, "I wonder what this place’s assessment program is like?" I suspect I am not alone in this.

My lack of curiosity about assessment when making an important choice about my children’s education probably surprises no one, but it should. It’s unsurprising in that no one, higher-ed insider or not, ever seems to worry about this when choosing a college. No admissions officer ever touted his institution’s assessment results. No parent ever exclaimed, "Suzy just got into Prestigious College X. I hear they are just nailing their student learning outcomes!" But it’s still a little surprising in that I am a professor and an administrator who has been involved in assessment in various forms for a long time. I have been dutifully doing assessment in my classes almost since I started teaching a decade and half ago.

Every year on my annual productivity report I write a mandatory and usually somewhat contrived narrative describing the ways in which I have changed my courses and teaching in response to the assessment data from the previous year. As an administrator, I sit on the Learning Outcomes Assessment Committee that oversees the institution’s assessment program and on the Graduate Council where we routinely critique new program and course proposals for the failings of their assessment plans.

So, what does it say that I looked at climbing walls, not assessments, when making a significant and expensive decision about my sons’ educations? It says that I, like virtually everyone else, don’t think that good assessment makes good universities and well-educated students or that bad assessment makes bad universities and poorly educated students. In fact, I am starting to wonder if assessment may actually do more harm than good.

What got me thinking about this was a New Yorker article by Atul Gawande on unnecessary medical testing and the low-value and sometimes harmful medical interventions that result from it. Drawing upon a number of recent studies, Gawande argues that much medical testing is unnecessary and that in addition to not providing useful information it can also lead to over diagnosis and over treatment. In one of his examples, he reports that ultrasound testing for thyroid cancer has made it possible to detect microcarcinomas that would have gone unnoticed before. These rarely pose a threat, but patients and surgeons find it difficult not to treat anything that sounds that scary. Gawande uses the example of Korea where thyroid cancer surgeries have increased drastically and thyroid cancer has become the most commonly treated form of the disease. However, mortality rates from thyroid cancer have not changed, while serious side effects from the all the surgeries have increased.

I saw unmistakable parallels to assessment in universities. Are we using assessment to find minor shortcomings in our teaching and curriculum, changing what we do in the hopes of remedying those shortcomings, and in the long run having no real positive effect on the quality of our graduates and institutions? Are we, in effect, finding and treating harmless academic microcarcinomas rather than real problems? And, if so, what might be the consequences of all this?

Has anyone looked into whether assessing student-learning outcomes over many years has made American colleges, or students, better in some way? Has anyone tried to compare institutions with different approaches to assessment? I am a historian so I am not familiar with the education research, but as best I can tell from a literature search and from asking people in the field the answer is "no."

To be fair, there is nothing directly comparable to mortality rates in higher education. Figuring out what makes one university better than another one or better than it was 10 years ago is tricky. But given the amount of time, effort, and money that goes into assessment, it would be helpful to have a track record of its efficacy.

Does assessment cause actual harm? Probably not in the way unnecessary medical treatment does, but there are opportunity costs associated with it. And most troubling of all is that the fundamental premise of assessment is that the problems we need to test for and try to fix are found in the classroom and the curriculum. So while we are agonizing about whether we need to change how we present the unit on cyclohexane because 45 percent of the students did not meet the learning outcome, budgets are being cut, students are working full-time jobs, and debt loads are growing.

People who work in assessment complain that faculty treat it as merely a compliance issue; that we just tick the boxes and don’t use the data to improve student learning. No doubt this it true. Advocates may be able to point to modest improvements in student learning in specific programs or courses with evidence generated by assessment instruments, but this is worryingly similar to surgeons patting themselves on the back for taking out tumors without checking to see if their interventions are affecting mortality rates.

If advocates could point to evidence that good assessment has led to improvements that are external to the process itself — like changes in a college’s reputation, ranking, or employment prospects for its students — I suspect faculty would give it more support.

Assessment is one of those things that we keep telling ourselves will pay off if we could just get it right, but we never seem to get there. It’s time for us to demand that the accreditors who are driving assessment provide evidence that it offers benefits commensurate with the expense that goes into it. We should no longer accept on faith or intuition that learning-outcomes assessment has positive and consequential effects on our institutions — or students.

Erik Gilbert is associate dean of the Graduate School and a professor of history at Arkansas State University.