Most college teachers assume that more tests are better than a few. Why? What caused us to decide on three or four unit tests followed by a final? Is there evidence that students don’t do as well in courses where there are only a midterm and a final? Why do we think that more tests might be better? And what do we mean by better? Higher grades? More learning?
In the article referenced below, authors True Kuo and Albert Simon review the literature on test frequency, and surprisingly, it is extensive. They rely heavily on a 1991 meta-analysis that compared the results of 40 studies analyzing how test frequency impacted student learning as measured by a cumulative final or standardized exam. The results (reported in the meta-analysis and in more recent research) are definitely mixed. In the meta-analysis, 13 of the 40 studies showed a moderate benefit to student learning for frequent as opposed to less frequent or no testing prior to the final. That means in the majority of the studies, no effect or a nonsignificant one was reported.
In addition to this rather surprising overall finding, there were other results of interest. When the results indicated a positive effect of frequent tests, the “student learning outcome … did not correlate with test frequency in a linear fashion.” (p. 157) This means that if two tests were beneficial, four tests were not twice as beneficial. “In other words, the test frequency effect diminishes as the absolute number of section tests increases.” (p. 157)
Then there’s the finding that when more tests and quizzes result in higher scores—scores on weekly exams tend to be higher than those on monthly exams, which makes sense because there is less material to study for each test—this improved performance on the more frequent tests did not result in better performance on the cumulative final.
Another finding relates to the role feedback plays in improving exam performance. Students learn more (as measured by exam scores) when each test is followed by a debrief session that focuses on their mastery of material missed on the exam. Authors Kuo and Simon say it is reasonable to hypothesize “that proper feedback and/or instruction has to accompany each test in order for the frequent testing to be effective in improving learning outcomes.” (p. 158)
Other evidence suggests that the test format needs to remain consistent throughout the course. The test frequency benefit is diminished when an instructor uses one kind of question and format on exams given during the course and another kind of question and format on the final.
One of the more consistent findings emerging out of this research is that student attitudes are more positive toward the course and instructor when they are given frequent exams. More tests and quizzes result in better attendance in class, and students find the exam experience less stressful when it occurs more regularly.
Reference: Kuo, T., and Simon, A. (2009). How many tests do we really need? College Teaching, 57 (3), 156-160.
Excerpted from How Many Tests? The Teaching Professor, vol. 23, no. 9, pg. 6.
