Tuesday, January 5, 2010

Groups: Random or self-selected?

This term, I'm organizing all my students in my lower-division ethics courses into learning groups. (For those of you who use the Blackboard CMS, the latest version makes it very easy to create such groups.) I plan to use the same groups throughout the quarter, for a wide variety of in-class and online activities (more on that later).

But the question I'd like to raise now is this: I used Blackboard's group building feature to place the students in randomly selected groups. I simply announced to students which groups they are in during the first meeting of the term. I think that randomly created groups are better, but I can imagine some reasons for permitting students to select their own groups.

Letting students select their own groups has some obvious advantages. First, if there are students in class who already know one another, then their groups are more likely to function well and have some initial momentum. Students might also feel that selecting their own groups is more respectful of their autonomy and maturity than being placed randomly in groups by their instructors.

But I opted for randomly selected groups, for what I think are decisive reasons. First, I need my groups to get organized quickly. I teach in ten week quarters, and if I let the students select the groups, they might take several weeks to do so. Second, random groups are likely to be more diverse than self-selected groups, in terms of gender, background, academic major, and the like. And it seems to me that a group benefits from the intellectual and cognitive diversity that may result. Lastly, being placed in random groups more closely resembles how students will later have to work in groups once they are employed. In workplaces, students may be asked to complete projects with people they have never met and who are fundamentally different from them culturally and in terms of their skills. So random groups prepare them for an important challenge later in life.

So: which is better — randomly selected learning groups or self-selected groups?


  1. I think that (quasi-) random groups are better, provided that the groups will work together on enough projects to get comfortable with one another and learn how to divide work effectively. If they're only working on one project together, even if it's a large one, I think it's better to let the students choose their groups.

    I agree with the reasons you gave for preferring random groups.

    I sometimes assigned groups quasi-randomly by sorting the students into freshman and non-freshman beforehand. They then counted off to form groups, ensuring that each group would have a mix of freshman and more experienced students.

    I once TAed for someone who had us collect some basic information on students (e.g., year, major, intellectual interests) on the first day. We then deliberately created diverse groups based on this information.

    One argument against randomly assigned groups: if the group work requires meeting outside of class, randomly created groups may have more difficulty coordinating schedules.

  2. Random.

    Students often choose classes with friends, eat with friends, study/hang out with friends. Being in a truly random group is an important experience for students, not only as preparation for the workplace, but also because meeting and working with different people should be a part of the college experience.

    Also, the positives for self-selecting seem oversold: groups of friends can also lead to group "drama," some students perform better without the distraction/expectations of their friends, students can be paralyzed by the new found autonomy (and responsibility) of choosing a group.

  3. Quasi-random

    I ask for students schedules and then try to organize groups in ways that will allow them to meet outside of class.

    Also, random can lead to students who have less than pleasant histories end up in the same group which can then be horrible. So, I also let students tell me, confidentially, if there's anyone in the class they really would have serious problems being grouped with. I know that outside of the classroom environment they'll have to interact with folks they may have a negative history with but (a) I don't know that I want my classroom to be the place where they work this out and (b) I've had too many situations where the drama of groups can so completely undermine my other learning goals that I figure it's better avoided when it can be.

    Of course, having ways to address inevitable drama as group work progresses is also important.

  4. I'm a big fan of random methods of selecting groups for in-class assignments. But, when students have to work together outside of class, you have to think about whether the random method will create more of a hassle for the students than it's worth. At a residential campus, there are few legitimate excuses for not being able to get together in the library, say. At a commuter campus, it might be better to let students figure out which persons have schedules that mesh well.

    By the way, you might want to be careful to distinguish between methods of assigning groups and resulting mixtures of abilities. One can construct heterogeneous groups; random methods may just as easily lead to homogeneous groups.

    Shameless plug: I have an article on just this topic coming out in the Spring issue of the APA teaching newsletter. Here's a copy:


  5. Another consideration for (quasi-) random groups is that there has been some research (which I can't put my finger on right now) which shows that minority group members are more likely to be socially excluded and perform less well when they find themselves the only minority in a working group. Perhaps this is also related to stereotype threat?

    Since my philosophy classes are 10-30% women, it would be likely for most randomly assigned groups to have just one woman. I readjust groups to prevent tokenism, even if that means some groups are all male. I imagine similar dynamics could hold for other sorts of minorities.


If you wish to use your name and don't have a blogger profile, please mark Name/URL in the list below. You can of course opt for Anonymous, but please keep in mind that multiple anonymous comments on a post are difficult to follow. Thanks!