Effective supervision and self – confidence : A balancing act

A- A A+

Effective feedback is crucial for career progression. Many women drop out of science, technology, engineering and mathematics (STEM) fields because they lack confidence in their abilities to succeed in these areas. One reason for this could be how supervisors give feedback. This column presents evidence suggesting that supervisors may hesitate to provide critical feedback to employees, particularly women with lower self-confidence. Instead, they might offer them overly positive feedback, hindering their performance in the long run. Getting the balance right between honesty and encouragement is vital.

Feedback plays a considerable role in driving high worker performance. Employees who work in different industries use the feedback they receive to improve their work and enhance their overall contribution to their organisation. So, it is important for this process to convey as much information as possible to the employee in question. While the nature of the feedback that employees receive matters, the frequency with which it conveys positive or negative information has a second-order effect: confidence.

How does confidence affect supervision and vice versa?

Consider an employer with two employees – Anne (she) and Bob (he) – who experiment with ideas for separate projects and implement one each. The employer would like both to produce successful projects. But without information on their ideas, both under-experiment. To solve this problem, the employer considers assigning a supervisor to Anne and Bob. The idea is that the supervisor would give them feedback to help them experiment more confidently. But the supervisor’s information is unverifiable, and so is the feedback he provides.

In this context, the supervisor may also need to be honest in the feedback he delivers. While useful in terms of encouraging employees to discard bad ideas, critical feedback can demoralise and discourage idea generation and effort. So, Anne and Bob’s confidence levels shape – and are shaped by – the effectiveness of the supervision they receive. The supervisor might feel more comfortable giving honest feedback if the employees are more confident. But if they’re not very confident, the supervisor might be more cautious about what they say, making supervision ineffective. It is a delicate balance.

In a workplace, feedback helps the employees improve their future performance. It also gives workers a signal about how smart or able they are and how likely it is to get a successful outcome given their ability. The more critical feedback they receive, the more convinced they become that they are not good enough and are less likely to succeed. For every project with different difficulty levels, employees will eventually stop trying new ideas, even if they can observe the quality of their performance. A supervisor who knows about this realises that a person with lower self-confidence will put less effort into implementing any idea he/she comes up with. So, when the employee’s self-confidence drops below a threshold, the supervisor stops giving them critical feedback. As a result, feedback becomes falsely encouraging. At this stage, the employee realises this fact and discards uninformative feedback. So, the employee, not knowing how well they are doing, will find no point in trying more and will stop experimenting, instead implementing the low-quality idea that they had in the first place.

This problem is more significant if an employee starts with very low confidence. When dealing with somebody with very low self-confidence, the supervisor, concerned about further discouraging the employee, would never begin to give any informative feedback, preventing the employee in question from even being able to start trying new (potentially better) ideas. This scenario of low self-confidence is where adding a supervisor is ineffective and would not improve worker performance.

For high self-confidence workers, supervision can be effective. When employees have high self-confidence, even without observing the quality, they believe that the likelihood of coming up with good ideas tomorrow is high, so they are willing to try, even without supervision. As a result, the supervisor knows that even if the employee is discouraged, he/she will try at least once without any informative feedback. In this case, it is less risky for the supervisor to send critical feedback. As a result, supervision becomes effective, improving employee performance.

While the supervision is effective in high self-confidence scenarios, a worker may enter into a low self-confidence state if he/she repeatedly receives negative feedback from the supervisor. So, the longer it takes for a person to succeed at their work, the more likely it is that the supervision will become ineffective (because the employee drops below the confidence threshold). In this case, it is arguable that it could be optimal for the supervisor to sometimes lie and encourage the employee in the high self-confidence region. Confidence is precious, it seems.

What about honesty?

When the supervisor is partially honest, the consequences are different. First, partially informative feedback cannot prevent the ineffectiveness of the supervision when it is delivered to a low self-confidence worker. Further, it causes early stopping even in high self-confidence cases. When the employee is unsure of his/her ability and the quality of his/her performance, feedback affects not only his/her self-confidence but also the belief about the chances of getting a high-quality idea versus the chances that a current idea is of high quality. This new information trade-off means that if the worker believes that it is more likely that today’s idea is of high quality (depending on the frequency of honesty) than the likelihood of tomorrow getting a good idea (depending on self-confidence), the worker will stop experimenting even if they are unsure of the quality.

Our findings are important for understanding the ways in which self-confidence affects how people respond to feedback. Firstly, the level of critical feedback someone receives can significantly affect their performance, especially for those with varying levels of confidence. People with higher self-confidence tend to receive more helpful feedback, which boosts their confidence further, enhancing their overall performance. This sheds light on why there may be performance gaps between men and women in fields driven by research. Studies have long shown that women tend to have lower confidence levels, which can hinder their performance, particularly in STEM areas. One reason for this could be that supervisors may not provide as much helpful feedback to less confident individuals, as highlighted in our latest work.

What should employers do?

The absence of informative feedback for less confident employees reflects an institutional problem with assigning a supervisor to motivate experimentation. But is there a viable solution to this problem? There are two possible solutions.

Confidence-building tests are one tool that employers can use. These tests need to be taken by the employees before they start working on a given project. We found that regardless of their initial confidence level, individuals value tests that provide more clarity about their abilities. Those with moderate confidence levels are also more willing to pay for such tests because they stand to gain more from increased confidence.

Our second solution suggests making projects more challenging, making it harder to develop good ideas. Surprisingly, this can lead to more effective supervision, especially for individuals with lower confidence levels. While more complex tasks may initially reduce the likelihood of success, they also make failure less indicative of ability, encouraging individuals to keep trying without feeling discouraged. This adverse effect of the project’s complexity makes it easier for supervisors to provide helpful feedback, particularly to those who lack confidence.

While these solutions are compelling in resoleving the problem of ineffective supervision for the employees with low self confidence, they can only be partially effective. The problem can only be fully resolved if the supervisor is involved in the implementation of the project. This raised questions about resourcing, and poses a further challenge to employers seeking to maximise the efficiency of their staff.


On a wider level, our work shows that feedback is crucial for fostering improvements in workers’ performance. Cultivating and protecting confidence is vital, for individuals’ own sense of self-worth as well as wider organisational productivity levels. For employers and managers, getting the balance right between honesty and encouragement is vital.

Author: Zeinab Aboutalebi, Ayush Pant