Harvard wimps out on testing

Mickey Muldoon

To oppose “results-based accountability” in education is close to a taboo nowadays, a position so antithetical to the spirit of the age that few dare mention it. Let us, therefore, declare ourselves shocked and saddened that Harvard University, in so many ways a pacesetter in education, is embracing that very position.

Starting in September, courses in Harvard’s Faculty of Arts and Sciences (FAS) will no longer routinely require “final” exams. For most of Harvard’s existence, any professor wishing to forego the practice of final exams required formal approval to do so by the entire faculty. At least since the 1940s, professors were required to submit a form to opt out of giving a final exam. But in fall 2010, professors will need to file a specific request to opt in. Dean of undergraduate education Jay M. Harris is already predicting reducing the academic calendar by a day or two in response to the eased testing burden.

Moreover, general exams—requiring seniors to demonstrate mastery of the fundamental knowledge of their major—are given in fewer and fewer departments. Even Harvard’s new General Education courses will abjure finals. We are left wondering: Without exams to prove it, how can students be sure that they are “generally educated” when they graduate? How can the institution itself be sure? Or doesn’t it care?

Some will say that other student work products—term papers, especially, but increasingly multimedia projects, too—are better gauges of learning than cumulative exams. Associate Dean Stephanie H. Kenen recently stated that “The literature on learning shows that hands-on activities can help some students learn and integrate the material better.”

In reality, however, the decline of testing at Harvard has little to do with any “literature on learning.” When we attended college there, more than four decades apart, some of our most fruitful educational learning experiences occurred in preparing for, and actually taking, final exams. They forced us to sharpen our thoughts and solidify our knowledge, whether it was by connecting the dots between Andy Warhol and Joseph Stalin for Louis Menand in 2006, or making sense of a year’s worth of American social history per Oscar Handlin in 1964. Term papers were essential, too—let us make no mistake. But they were easier to fudge with obscure research, borrowed insights, and artful prose. It was finals that forced us to think, to synthesize, to study, and to learn.

What’s really happening, we sense, is that Harvard is yielding to education’s most primitive temptation: lowering standards and waiving measurements for the sake of convenience. It certainly isn’t the only university to succumb, but given Harvard’s reputation as a trendsetter, we should expect better. Just imagine: Students will be delighted to forego finals and instructors will be thrilled not to have to create or grade them. Everybody finishes the semester earlier. (The last few weeks of class don’t really count when that material won’t be tested!) Yet Harvard’s leaders may eventually have to acknowledge that, with fewer exams and test results, they will know less and less about what students are or are not learning within their hallowed gates.

Not so long ago, Harvard was striding toward transparency and accountability. In 2006, under the leadership of interim President Derek Bok (no slouch himself as an education reformer and critic), the university participated in the Collegiate Learning Assessment (CLA). The CLA is intended to measure the kinds of skills and thinking at the core of an “arts and sciences” curriculum and, by comparing the scores of seniors and freshmen, to gauge a university’s “value-added.” One of us, a senior at the time, even volunteered to participate. It was a rare chance to put the old joke to the test: “Why is Harvard such a great repository of knowledge? Because students enter with so much and leave with so little.”

Sadly, Harvard’s CLA results were never shared with participants, as had been promised, much less with the outside world. The flickering light of results-based accountability at Harvard was thus dimmed—by whom and why, we can only guess.*

Granted, testing is complicated. How to assess a semester’s worth of learning in 180 minutes? How to probe what one has learned during three years as a history major? How, simultaneously, to measure the accumulation of knowledge and the development of analytic skills and effective expression? How to distill course themes into challenging essay questions or problem sets, and how to grade them fairly?

But avoiding tough methodological challenges isn’t in Harvard’s mission statement. In matters of education policy—including many earlier rounds of assessment, such as the SAT and Advanced Placement exams—Harvard has long been a pioneer. Other universities look to it for guidance. Why not with final-exam style assessments, too? Harvard’s Graduate School of Education is full of testing experts and its Psychology Department is stacked with heavyweights. Its mathematicians, computer scientists, and statisticians are competent to crunch the numbers, sample the populations, etc. Couldn’t they help the university develop suitable guidelines, templates, and prototypes for measuring what their students learn? Would it be too much to ask them to actually develop better tests? Maybe even share them with the world?

Harvard might fruitfully take a cue from K-12 education. Here we’re seeing slow but steady progress towards intelligent assessment and fair accountability. The primary-secondary education community is approaching consensus on content standards for math and English language arts. Consortia of states are undertaking the development of “next-generation” assessment systems. The Obama Administration has taken stock of No Child Left Behind and offered a new blueprint for giving schools and districts more flexibility to reach higher performance standards. None of this has been easy, and countless political headaches would have been avoided by simply jettisoning results-based accountability. Plenty of teachers would have been pleased, too. But most K-12 policymakers know better: Were it not for the dreaded tests, we would not be able to learn from our educational successes, nor direct attention to our most persistent failures.

Harvard doubtless assumes that no formal measures of learning are needed to demonstrate its educational value to students. Just peek inside Lamont Library late on a weeknight and behold the heaps of books, index cards, and coffee mugs. Listen to the keyboard clatter of great term papers in the making. Well, we studied in Lamont—one of us quite recently—and we have a secret to share: There is a difference between effort and learning, between putting in the time and coming out with something worthwhile. For every undergraduate writing the next Great American Novel, another student is frustrated, confused, and stressed by ambiguous expectations from instructors. Harvard’s struggles with mediocre instruction would be greatly aided by better tests aligned with clearer expectations—not by giving up on exams altogether.

Harvard is blessed with talented students—it can pick and choose among America’s finest—and that doubtless encourages it to pay scant attention to how much they actually learn during their undergraduate years in Cambridge. University leaders also understand that public accountability can be humbling. Arrogance and pandering are more convenient. They just don’t get us any closer to veritas.

*The authors contacted the office of the President last Friday to corroborate this account of the CLA at Harvard. As of Wednesday, July 14, officials were able to neither confirm nor deny it.

By Chester E. Finn, Jr. and Mickey Muldoon

Finn and Muldoon both graduated from Harvard, classes of 1965 and 2007, respectively.

In this video, Mike Petrilli interviews Finn and Muldoon about the future of higher ed testing and accountability.

More By Author