Best Group Learning Experience - Step 1: "Who shall I work with?"

Group work is almost unavoidable in the lab. How the groups are formed has a markedly big impact on how effective the group learning experience. As pointed out in  "Better Group Work Experiences Begin with How the Groups are formed", there are three common ways of forming groups, randomly formed groups, teacher-formed groups, and student-formed groups. The timing is just right -I just had experience using all three approaches, and started to survey our students which is the best learning experience. I'd like the share my thinking of these three approaches here.

Randomly formed group - in ASP201/202 lab (University Physics), we randomly assign the groups. The group activities include hands-on exploration in the week, and an in-lab group assignment (discussion + questions) each week. The advantages are obvious - students are able to know everyone in the section and they quickly figured out who they like to work with. In the second semester, when we had transfer students,who were obviously left behind, and unable to contribute to the group work, the high-performing students started to complain.

Teacher formed group - in the last four weeks, CHE142 (Principles of Inorganic Chemistry) Lab, as the project requires 4 students instead of 2 students, we re-arrange the student-formed pairs, and assign them into group of four. The group activities include hands-on exploration and a weekly homework.  The strategy that we used is to mix the high-performing and low-performing students together, hoping the high-performing students can help and lead the team. The majority of the groups reported minor even major problems - high-performing students complaining that they had to do most of the work otherwise they lost points from the group assessment; the low-performing students complained that they were not trusted, and were not allowed to contribute to the group project. 

Student formed group - I changed the approach this semester and asked physics lab students to choose their own partners and informed them that, the groups don't have to be permanent and I am open to their suggestions. Students group with their friends or who they know better. New transfer students tend to just group with who they stayed the closest in the first class. At the midpoint of the semester, students were surveyed anonymously if they are satisfied with their current group, all of them have positively expressed that they'd like to stay in current group. 

Although there's no one-approach-fits-all, right now, student formed group works the best for me. 

1) First of all, it allows students to find who they like to work with, or who they know they can work with. Group work is not new to many students and students know each other from many group activities in other classes, or in campus wide events. If they get along well, they would likely want to collaborate again. It takes time for students to get to know each other, and get to know how to collaborate with a particular person, if all the past experience has shown them who the best partner is, there's no reason why we don't take advantage of that. As I wrote in my recent journal (reflecting on power of introverts), having the authority to choose partners would encourage introverts students to participate in a comfortable small group setting, and they are more likely to engage in the discussions. 

2) Secondly, it empowers our students, and encourages them to take ownership of their learning. Working with friends asks our students to transform the social relationship into task-oriented relationship in group activities. Honestly, it doesn't work all the times. However in a student-formed group, students are more likely to seek ways to make it work (instead of complaining why the teacher assigned them to be with particular people). They tend to be more patient and understanding when there is a problem in group work. A group needs both leaders and members, and needs a collection of experiences, skills, knowledge and abilities. Some teachers will survey the students with past experience, and incorporate the data into their decision of assigning groups. However, how accurate is this approach? Can we guarantee that the theoretically correct groups will work the way they are supposed to work?  Not even need to mention the time and efforts spent in surveying and data collection. If I am not confident of my decision that we'll make, I'd rather have the students naturally choose who they'd like to work with, and be willing to solve problems. 

It is also worthwhile informing the students that, the groups are not permanent. It is possible to request regrouping if necessary. I also found this post "Students Riding on Coattails during Group Work? Five Simple Ideas to Try" useful when designing group projects.