Feedback: can we learn to calibrate our confidence?
An employee may confidently predict a deadline and later discover that it was unrealistic, without reconsidering how they assess uncertainty. The managerial question is specific: what kind of feedback helps bring expressed confidence into line with the actual reliability of forecasts?
What the literature establishes
Sarah Lichtenstein and Baruch Fischhoff (1980, Organizational Behavior and Human Performance) studied calibration training: learning to align stated confidence with the frequency of correct answers. In their judgment tasks, training that included feedback on this relationship improved calibration. This finding supports the possibility of learning, not a promise that an ordinary managerial comment will produce the same effect.
Confidence is not a single phenomenon
Don A. Moore and Paul J. Healy (2008, Psychological Review) distinguish three forms of overconfidence: overestimating one's performance, placing oneself too high relative to others, and assigning excessive precision to one's beliefs. These phenomena do not necessarily move together. Telling someone they are “too confident” therefore identifies neither the observed problem nor what they need to learn to correct.
The premature conclusion
One might conclude that good feedback should make employees more cautious. That would confuse calibration with modesty: people can also underestimate the reliability of their judgments. The aim is not to reduce confidence systematically, but to make its different levels more informative. (our executive and employee training programmes)
What the evidence and current practice leave unresolved
Experimental tasks have verifiable answers; workplace forecasts often concern events whose definition or deadline changes. Moreover, a handful of forecasts cannot establish calibration, and a single outcome cannot establish whether a probability was right or wrong. Without a record of confidence expressed before the event, feedback risks merely reconstructing, with hindsight, what everyone should have known.
A practical check in Berne
As part of SHR's “L'art du feedback” programme, an exercise could involve recording binary forecasts before their outcomes are known, each with a probability, a deadline and a verification criterion: a file processed on time in the federal administration, a milestone reached in public health, service restored in telecommunications, or a batch accepted in precision manufacturing. For each probability band defined in advance, the measure would be the gap between the average predicted probability and the observed frequency of occurrence, reported alongside the number of forecasts. Comparing these gaps across successive sets of comparable cases would give feedback a verifiable focus in Berne, without automatically attributing any improvement to the training. To go further: explore the The Art of Feedback training in Bern in the canton of Bern, or browse our executive and employee training programmes in Switzerland.
In pictures: The Art of Feedback in Bern
- the art of feedback
- Bern
- research
- training bern