Via Joanne Jacobs, here's an interesting research tidbit—highly preliminary and tentative, but still interesting. A couple of psychology professors at the University of Texas started giving students in their intro lecture course a brief online quiz in every single class session. They found that average grades went up modestly, both in their class and in other classes, though this was tricky to assess since previous classes had used different grading curves. However, the daily quizzes did unquestionably improve the relative performance of students from low-income homes:
There's really not enough data from this one study to figure out why the delta between high and low-SES groups compressed with daily testing, but the researchers' best guess is that the low-SES students benefited more from the daily, immediate feedback:
In our view, the patterns of improved performance across three outcomes (in Introductory Psychology, in other Fall classes, and in subsequent Spring classes) most plausibly reflect changes in students’ self-regulated learning — their ability to study and learn more effectively....In particular, students had to adopt reading, note-taking, and study habits that allowed them to keep up with the material. In talking with students, many noted how they had learned to set aside specific times to prepare for each class–something that they did not initially feel they needed to do for other classes. The repeated testing also broke the material into segments that required students to focus their attention on the relevant content and the immediate feedback after each quiz provided students with a constant and objective means with which to engage in productive self-evaluation. The daily quizzes also encouraged students to attend classes at higher rates.
In other words, the high-SES students had better average study habits to begin with, so the daily testing affected them only modestly. The low-SES students had poor study habits, and the daily testing made them face up to this early in their college careers and do something about it before it spiraled out of control. This affected not just their performance in the psychology class itself, but in the rest of their classes as well.
There are obviously a ton of confounding factors that could be at play here, but it's an interesting result, well worth following up on.