It Just Takes One Bad Apple to Disrupt a Team

We recently wrote an important paper focused on education for collective intelligence (CI). We argue that CI needs to be developed as a core skill throughout the lifespan, and that the most important driver of CI skill development is education. Groups working together well can achieve more than individuals working alone—organisations recognise this, and employers are increasingly looking for graduates who have an ability to work as part of a team.

But our education system does not generally cultivate CI or prepare students very well for teamwork. Education is primarily focused on individual learning outcomes. Also, in practice, the reality is that teamwork is difficult for educators to coordinate, and students are often averse to teamwork. Teamwork is hard work much of the time; it takes a lot of practice and sometimes it’s deeply unpleasant. There are lots of things that can and do go wrong. For example, as one saying goes, all it takes is one bad apple to ruin the experience for everyone.

Examination of the "Bad-Apple" Effect

In a recent study, Chapman and Meuter (2023) examined this "bad-apple" effect. They examined the experience of 409 business students working across 105 groups. In the courses that students were taking, everyone understood that teamwork was critical for project assignments, and they understood that peer evaluations would feed into their final project grades. Using an online peer-evaluation tool—the Comprehensive Assessment of Team Member Effectiveness (CATME)—students rated themselves and each of their teammates on indicators of team effectiveness, including their contribution to teamwork, interactions with other teammates, keeping the group on track, their expectations regarding quality, and their knowledge, skills, and abilities.

Chapman and Meuter also measured key indicators of team dynamics, which students provided ratings on, including team conflict (i.e., task conflict, relationship conflict, and process conflict), interpersonal relationships, team cohesiveness (i.e., task attraction, interpersonal cohesiveness, and task commitment), and team satisfaction. And, finally, as part of project evaluations, teams received grades from their instructors on different project deliverables: the project proposal, the data collection tool, and the final report.

To identify teams that included a "bad apple," Chapman and Meuter used a simple algorithm derived from the CATME peer-evaluation data. In particular, the "bad apples" were students who were collectively rated by their peers as 20 percent below the team average on key indicators of effectiveness. Although teams with a bad apple had similar average GPA scores when compared with groups that had no bad apple, Chapman and Meuter identified a significant issue when it came to team project performance.

Performance and Group Dynamics

Specifically, aggregating project grades across all three project deliverables, it was found that 21.3 percent of students earned an A grade if they were in a group with no bad apple, whereas only 10.0 percent of students earned an A grade if there was bad apple in their group. Furthermore, having a bad apple in the group had a negative effect on group dynamics: These groups had lower overall team member effectiveness ratings, lower interpersonal relationship scores, higher team conflict, and lower team satisfaction. Finally, the group dynamic ratings of both interpersonal relationships and team cohesion predicted project grades. The findings are quite stark: One bad apple in a group had quite a negative effect on group dynamics and team performance.

Chapman and Meuter were also able to distinguish between groups that used single peer evaluation (i.e., one evaluation at the end of the course) and groups that used a two-stage peer evaluation process (i.e., with one peer evaluation delivered as feedback after the first project deliverable, and the second peer evaluation delivered at the end of the course). Interestingly, the two-stage peer-evaluation groups had a lower prevalence of bad apples at the end of the course (9.9 percent as opposed to 16.1 percent in the single peer-evaluation groups). Also, compared to the single peer-evaluation condition, groups using a two-stage peer-evaluation process had less team conflict and higher ratings of team satisfaction at the end of the course. One student in the two-stage peer-evaluation condition summarised the basic group dynamic at play: “Everyone in the group really stepped it up after the [first] CATME survey.”

Addressing the Problem

The bad-apple effect is a real problem, and in my experience working with student teams, they intuitively recognise this as a potential problem. In addition to providing early feedback to group members that may act as a corrective to ineffective member and group dynamics, Chapman and Meuter suggest that student teams may benefit from training in team basics (e.g., use of agendas, taking meeting minutes) and other aspects of effective teamwork.

Also, given how busy teachers are, Chapman and Meuter suggest that educational institutes seeking to cultivate teamwork skills in students might benefit from the use of team coaches. This is similar to a suggestion we make in our CI paper, in which we recommend including group process facilitators as an adjunct role supporting teachers in the classroom, particularly when students are working on complex CI and teamwork design projects. Historically, it is evident that educational technology and system designs have been modified in response to environmental challenges—and this process of adaptation will likely continue into the future. If we can figure out how to design good educational systems supporting teamwork and CI, I think we’ll see fewer "bad apples" as more students develop enduring teamwork skills that are further refined and passed on to future generations.