Participant Selection and Group Formation

The selection process for your event depends on the desired outcomes of what you are asking participants to do. At MacChangers, part of our goal is for teams to work in multidisciplinary and multilevel groups. This means that groups are made up of students from all levels of study and from all programs at McMaster, including post-graduate degrees. Often, people trained in different fields and at differing stages in their studies have diverging views on topics and problems they may encounter. By allowing for multidisciplinary, multilevel groups to be formed, we feel that students develop unique solutions and learn new skills and abilities from one another.

Team Size

 At MacChangers, our ideal group size is four students. Groups of either 4 or 5 allow for varying perspectives and opinions to be expressed while being small enough to maintain unity and cohesion between group members. We also find that teams of 4 or 5 are resilient to a group member leaving the program without putting too large of a burden on the remaining participants, as may happen with groups that begin smaller. As such, by the end of our program teams typically end with 3 or 4 members. On rare occasions, groups of 2 have completed the program.

Forming successful groups is a bit of an art form and as an organization we are continuing to update and change our process year by year. It is important to both try to create a functioning team dynamic, as well as a team that we can logistically put together given our pool of participants. Anecdotally, one way to sink a ship in a collaborative program is to have a dysfunctional team. Below we share our current process for accepting participants and forming teams; however, this is subject to change and an area we are continuously interested in improving.

Applicant Selection Process

Our selection process involves a series of both close-ended and open-ended questions.

Our typical admission form includes the following:

• Asks for name and preferred name, pronouns, contact information, program, and level of study.
• Asks participants to confirm that they understand their time commitment to the program and to confirm they understand the program outcomes.
• Selection of the theme area or topic.
• Determine whether they have entered alone, with a friend, or as a group of four.
• Expression of interest.
• How they heard of our program.

Theme selection

In our program, students will each participate in 1 of our 4 theme areas. This is based on interest and space available in the program. Our theme selection process has proven to be very helpful. Applicants fill out a form in which they rank each possible theme or topic into three categories:

1. I am not willing to work on this topic.
2. I am willing to work on this topic if my preferred choice is unavailable.
3. This is my preferred topic.

Applicants are free to rank each topic however they please, with the acknowledgement that students who rank topics in the second or third category have a greater chance of being selected for the program. We’ve found that students have fewer conflicts when beginning their projects when they are happy with their theme area. We propose that this might help to reduce attrition due to conflict caused by disinterest in their challenge area. That said, we’ve yet to study this effect formally.

Open-Ended Question

We use the expression of interest to determine the enthusiasm of participants in the program.  Using at least two reviewers, we review their understanding of the program goals and their motivation for applying. Applicants with poor responses are rejected while participants with effortful responses are accepted. For example, if an applicant says something that is factually incorrect about the program they will be rejected. Individuals who can speak to personal goals and an accurate understanding of the program are generally accepted. We typically generate a third section of applicants who are neither immediately rejected nor accepted. If the pool is competitive, we typically are unable to take these applicants; however, we keep them on a waitlist should any successful applicants later decline the program offering.

Using open-ended questions for participant selection is not recommended for programs that will be multilingual or accepting applicants from universities from various countries. This question may put those who are not native speakers of the language in which the question is asked at a disadvantage as they may be unable to express themselves appropriately. This question is beneficial when all applicants are level in terms of ability to express themselves in a language.

Use of GPA

You may notice that we do not ask participants their grades or GPA. This is an intentional choice; our program goal is meant to inspire students and to help them feel encouraged with experiences outside of the traditional classroom. Further, our program is co-curricular, and a commitment that students choose to make outside of their course load. We do not want to make our program feel like it has the same pressure as their already heavy course and assignment schedules. That being said, GPA can be a good tool to use when forming groups. Literature suggests that student groups in which participants are all of moderate ability or GPA are better suited to be more uniformly grouped. However, groups of combined high and low ability or GPA tend to have positive group dynamics and are better in diverse groupings (Chen & Kuo, 2019). GPAs can also be used for cut-off in competitive programs; however, studies do show that GPA variety tends to help as teams will be able to share more diverse knowledge more effectively than if they were all of the same ability level (Webb, 1982; Chen & Kuo, 2019).

Group Formation Process 

The group formation process begins by removing students who fail to demonstrate their enthusiasm or understanding from the successful applicant pool. At MacChangers, we pre-determine the number of applicants that will be selected based on the size of our staff. Our program relies on each team being paired with a peer mentor and we try to ensure this number is a ratio of five teams of four students, to one mentor.

Successful applicants are divided into their most preferred theme areas, then sub-divided into their program, program year, and type of response submitted for the situational question. We attempt to diversify the groups as much as possible.

We do allow participants to enter as an existing group of 4 or as a group of 2; however, all participants of the group or duo must be accepted. Pre-formed groups do not work for every program, however, given our program is co-curricular, we try to accommodate students whenever possible. This can defy our goals of creating multi-disciplinary teams and may be removed as an option in the future. An alternative option to remove this issue, is only allowing for pre-formed groups of 2 to join, thus allowing for the rest of their group to be diversified. Familiarity within a group seems to improve the comfort students have in sharing information effectively, however, a group with high familiarity is less likely to have differing knowledge and perspectives (Chen & Kuo, 2019). This means that only allowing pre-formed groups of 2 may improve the balance between familiarity and diversity.

That being said, we are currently undergoing research on project continuity and what allows best for a group to be able to continue with their project. We find that within our program, groups that are pre-formed tend to fare better with the continuity of their projects. It appears that groups with high internal familiarity, who are preformed, will have better group dynamics and want to continue together. So, arguments for and against preformed groups can be made, and the decision may be reliant on your own program’s goals.

Current Exploration and Research

It is important for programs to think about how they are forming their groups, and the above can be used as a guide for how we have thus far been successful. The following is a discussion of our ongoing research into ways that we can improve, and potential jumping off points for the future of our program.

A large focus of our research, and of our current method of group formation, is the importance of both heterogeneity and homogeneity. Heterogeneity in a group setting refers to the differences found among groups and the push to make them diverse.  Homogeneity in groups refers to the aspects that should be shared among individuals within or between groups. Studies have identified that the most advantageous groupings are those that are inter-homogenous and intra-heterogeneous, meaning that groups formed are similar to one another to minimize advantages, but internally diverse so that within a group there are varying perspectives and traits (Chen & Kuo, 2019).

Heterogeneous groups are found to outperform homogenous groups on a wider variety of tasks. Moreover, heterogeneous teams outperform student-formed and randomly formed teams as well (Chen & Kuo, 2019). This just provides further reason for minimizing the number of people allowed in a pre-formed group, and the need for a standardized approach to assessing student traits.

As alluded to above, we are reconsidering how we form groups and diversify individuals. We are interested to see if implementing a research-tested system into our existing application process would improve the validity and efficacy of our group formation process, to reduce attrition and increase cohesion within groups. An exciting area in group formation and social networks is the use of algorithms to form groups. However, this is not feasible for the staff and capacity of MacChangers program. While this approach may work for other programs, given our small participant pool, it seems more feasible to utilize a more questionnaire-based approach that can be used to loosely categorize individuals. This could then be used in conjunction with our pre-existing techniques to create diverse groups. The following are approaches being considered for future exploration.

Felder Silverman Model

In this approach learners are distinguished on a basis of 4 dimensions; active or reflective, sensing or intuitive, visual or verbal, and sequential or global. On the first dimension, active learners prefer to really engage with learning material and be able to apply it, while reflective learners prefer to work in smaller groups and to reflect on material. Sensing learners prefer to learn concrete information and facts and they may be more realistic, whereas intuitive learners enjoy abstract information, theories and being creative. On the third dimension, visual learners tend to prefer learning from their observations, while verbal learners gain more from textual representations or speaking. Finally, sequential learners learn in steps that have a linear process, where global learners like to absorb learning material all at once and without finding connections but after learning it all they will make connections within the big picture. (Graf, Viola, Leo, & Kinshuk, 2007)

To find where students lie on these dimensions, Felder and Silverman identified 44 questions, with 11 questions pertaining to each dimension. This test is known as the Index of Learning Styles (ILS). This has been found to be reliable, and the authors of an analysis of this method have compiled a list of only 20 questions that are the most representative (Graf, Viola, Leo, & Kinshuk, 2007). This method would potentially be effective as while 44 questions may be daunting to pose to applicants, 20 is far more reasonable. Questions can even be further reduced by selecting only questions on dimensions that pertain the most to your program.

Where students stand on these dimensions can then be utilized to form diverse or heterogeneous groupings, where for example groups are formed with both visual and verbal learners, as well as both sequential and global learners. This may improve students’ abilities to take on projects from a variety of perspectives as well as to learn from one another’s differing abilities.

Kolb Questionnaire

The Kolb Learning Style Model defines an individual’s learning based on how they perceive and process information. Learners can perceive information as either concrete experience (CE) or abstract conceptualization (AC). CE learners are peer oriented while AC learners are symbol oriented and learn best under the direction of authority. Learners can also process information through either active experimentation (AE) or reflective observation (RO). AE learners prefer to “do” and learn via experimentation, while RO learners must rely on observations to come to conclusions (Lu, Jia, Gong, & Clark, 2007).

An analysis of Kolb Learning Styles and online learning identified 4 groups of learners based on how they perceived and processed information; Divergers (CE & RO), Assimilators (AC & RO), Convergers (AC &AE), and Accommodators (CE & AE). The test utilized here is the Kolb’s Learning Style Inventory. This method has been refined to improve validity and can now be considered a useful tool in determining learning styles (Lu, Jia, Gong, & Clark, 2007).

Once again, this method could be used to form groups of 4 composed of at least one individual from each learner group (Diverger, Assimilator, Converger and Accommodator). This may pose an issue as an equal amount of each learner type may not be available in a participant pool, but it is possible that simply aiming for as much heterogeneity as possible would improve group dynamics.

Big 5 Questionnaire

The Big 5 Questionnaire is a method that relies on the Big 5 personality traits: Openness, Conscientiousness, Extraversion, Agreeableness, and Neuroticism. If an individual is high in openness, they are intellectually curious and prefer variety, high in conscientiousness is an individual who is disciplined and achievement driven, high extraversion indicates high sociability and assertiveness, high agreeableness is a helpful, cooperative and sympathetic learner, and a highly neurotic individual has anxiety and self-doubt. An analysis of this method identified the importance of personality in conjunction with learning styles, even finding that high openness had correlations to higher GPAs (Komarraju, Karau, Schmeck, & Avdic, 2011).

These traits may be assessed utilizing the Five Factor Inventory (NEO-FFI) which utilized 60 questions. However, a revised version, the NEO-FFI-R was found to be more reliable and consistent (McCrae & Costa Jr, 2004). This test, while revised, is extensive and 60 questions may not be feasible for all programs. It would lend to the ability to create heterogeneously formed groupings of individuals who differ on the 5 traits.

Conclusions

The major findings on group formation are that heterogeneous groups are the most effective in terms of group collaboration, efficacy of knowledge sharing, and success of the group. It is also important to ensure that groups are small enough to feel comfortable sharing with one another. MacChangers is taking steps to increase the diversity within groups, by assessing our choice to allow pre-formed groups, diversifying students based on level and area of study, and by assessing the validity of our current methods. Going forward, best practices seem to rely on the use of a standardized method of assessing participant traits and considering learning styles in how groups are formed to ensure lower levels of attrition and higher rates of group success. The above are guidelines that can be used and altered for other programs that have been found effective in MacChangers, as well as avenues we are continuing to learn about and explore.