King's College London

It is well-established that people hold ‘self-servingly’ biased beliefs about their own personal traits and abilities that help bolster their self-esteem.

A new study, published in the Economic Journal, has tested whether this phenomenon is linked to how we assess others.

The study looked at the findings of an experiment involving 426 students. In the first part of the experiment they were asked to participate in a computerised quiz. In the second part they would be paired with either a human partner who completed the same quiz, or with a virtual robot. They received feedback on their team’s performance and asked to assess their own and their partner’s performance.

In the third part of the experiment participants were given an opportunity to swap to a different, randomly selected partner before assessing their performance again, just as in the second part.

The researchers found that when they were partnered with a person, participants tended to demonstrate self-serving biases. They gave more weight to positive than to negative feedback when assessing their own and their partner’s abilities and skills. This resulted in participants being more confident in their performance.

But it also caused more mistakes as they turned out to be more hesitant to switch partners in part 3, even if those could have been potentially better skilled.

By contrast, when they were partnered with a virtual robot, the participants took positive and negative feedback similarly into account when assessing their own performance and the robot’s skill. As a result, they held more accurate beliefs about their performance.

The research findings have a number of practical uses in the workplace, to help improve feedback and development for staff.

Co-author, Dr Zahra Murad from the University of Portsmouth’s School of Accounting, Economics and Finance, said: “The behaviour observed in the controlled settings of a lab can manifest in various aspects of our lives, from academic group projects to professional teamwork. You might find yourself overestimating a colleague's proficiency to justify delegating tasks, thereby avoiding the stress of extra work and the potential revelation of your own shortcomings.”

Co-author, Dr Alexander Coutts from Schulich School of Business, York University, added: “Consider a scenario where you want to see yourself as the key player in a team project. The simple solution for ensuring success might seem to be monopolising the workload.

“However, this approach comes with its risks—overburdening yourself or facing the uncomfortable truth that you might not be the infallible expert you thought. Instead, our minds seek a simpler solution by reevaluating our partner’s capabilities.”

With regards to employee feedback and development opportunities, the research suggests that organisations should be cautious when solely relying on team performance feedback.

Dr Zahra Murad added: “Overconfidence seems to be less of a problem when employees co-produce with a robot or an AI tool. In such situations, individuals exhibit a greater capacity to integrate performance feedback and gain insights into their genuine abilities.”

The study is a collaboration between the University of Portsmouth and King’s College London in England, and York University in Canada.