Requiem for the final exam?

The Boston Globe reports that university faculty are giving final exams less frequently:

In the spring term at Harvard last year, only 259 of the 1,137 undergraduate courses had a scheduled final exam, the lowest number since 2002, according to Jay M. Harris, the dean of undergraduate education. Harris said he’s hesitant to read too much into the numbers, which, he said, don’t include whatever final exams were scheduled in language courses, don’t reflect the other forms of assessment that have replaced exams, and don’t account for small seminar classes, which typically would not have a traditional, sit-down, blue-book final.

Now I don't know if this trend extrapolates to the non-elite colleges and universities, but it's certainly worth thinking about from a learning-oriented perspective. (Note that ISW has had quite a few discussions of exams: their merits, how to design them, etc.)  One reason to think this is a welcome trend is research suggesting that more frequent evaluation, focusing on smaller learning increments, leads to greater learning and mastery. But I have to say I'm still attracted to the idea that a final exam compels students to think synthetically and comprehensively about what they've learned, thus helping students to reinforce what they've learned and appreciate both its scope and value. Here's Robert Bangert-Drowns, dean of the school of education at the University at Albany SUNY:

There’s nothing magical about finals, Bangert-Drowns added. They can be arbitrary and abstract — an inauthentic gauge of what someone knows. Research, by Bangert-Drowns and others, shows that frequent testing is more beneficial. And yet, many still find value in the final exam. It might be stressful, even terrifying, but it has the singular power to force students to go back over material, think critically about what they have read, review hard-to-grasp-topics once more, and even talk about the subject matter with classmates and instructors — all of which enhance learning.

 And finals seem to provide a useful summative or chronological measure of student learning.

“You can measure an institution’s performance, a department’s performance,” said James Engell, Gurney professor of English literature and professor of comparative literature at Harvard. “But the real question is: How much did your students learn? How much better are they at something now than they were when they started? And I think examinations — whether they’re final examinations or other kinds of examinations — play a real part in that.”

In the end, my suspicion is that 'frequent small evaluations' vs. 'occasional big evaluations' is a false dichotomy. Some confirmation of this:

One ... study, published last year, focused on more than 1,500 students taking algebra at Richard J. Daley College in Chicago between 2004 and 2006. M. Vali Siadat, the chairman of the math department there, compared the outcomes of algebra students who took weekly, cumulative quizzes over the course of the semester with those who received less rigorous, regular assessment.Those tested weekly not only did better on the midterm and the final exams, but better overall, outperforming their classmates who did not receive regular quizzing by about 16 percent by the end of the semester. With regular, cumulative testing, Siadat concluded, the students were simply better prepared.

“The students know the final is just another act,” said Siadat, who coauthored a paper on these findings last year with his Richard J. Daley colleague Euguenia Peterson. “It’s going to be cumulative just like the previous tests. They’re ready for it and they tend to do well.”

My own practices tend to integrate these two approaches. Many of my classes have weekly online quizzes or short writing assignments. For the last five courses I've taught, I had a final in four out of five (it was optional in one course).  But in each course, I've tried to limit its impact on students' term grades, as well as exam-related anxiety, by making the final 20-25% of the overall grade.

I'm hoping I'm using evaluation in a way that really is conducive to student learning, but I'm interested in whether others note the trend in the Globe story: Are you giving final exams, and how do you see their role in helping students to learn philosophy?