Let's look at some project ideas that didn't work out well.

You should form groups [link to forming a group discussion] and start thinking about project topics right away and get timely advice from the instructor. One semester the projects did not go well because almost all the groups of students did surveys [link to definition of surveys]. It would be fine if some groups did this sort of project, but a survey is a limited research tool if it does not involve any experimentation.

For example, in one typical project, a pair of students interviewed about 100 students and collected data on their age, family background, and marriage plans (whether they expected to marry and, if so, at what age). The analysis included various scatterplots, correlations, and two-sample comparisons. Even if this sort of project is done very well, it is unlikely to yield interesting results. In addition, there was a sameness to seeing survey after survey. Even in the context of observational studies, the rest of the class would find it more interesting to see a variety of study topics, for example animals, the news media, raindrops, or just about anything different.
It often takes some discussion [link to "how to discuss project topic as a group] to isolate the fundamental question underlying a research idea [link to discussion on research idea and question]. For example, students often have the idea of gathering from students data on grade point average (GPA) and study habits, or alcohol consumption and grade point average, or number of hours studied and amount of time spent each week exercising. Students have a natural interest in questions like the effectiveness of studying, differences between athletes and non-athletes, and campus drinking. However, these sorts of questions rarely yield interesting results. Correlations between the measurements are usually weak and, in any case, it is not clear how to interpret the associations. With this in mind, we intercept these projects at the beginning stage, by excluding them from eligible project topics [link to suggested project lists].