How Confidence Spreads in the Workplace and Beyond
Every workplace has its own acceptable volume of certainty. In one team, saying “I think this is right” sounds appropriately careful. In another, the same sentence is heard as weakness, while the person who speaks without qualification becomes the centre of gravity. This dynamic influences not only how people communicate but also how they internally judge their own decisions.
We often adjust how certain we sound to persuade others, avoid embarrassment, or fit the room’s culture. But a new study published in Current Biology raises a more unsettling question: after matching someone else’s confidence in public, do we carry a changed sense of certainty back into our private decisions? According to researchers Einar Andreassen, Chris Frith, and Daniel Yon from Birkbeck, University of London, the answer appears to be yes.
Their study found that participants paired with a confident decision partner became more confident themselves, while those paired with an uncertain partner became less confident. Remarkably, this effect persisted even after the partner disappeared and participants returned to deciding alone. This suggests that confidence is not merely a public performance but can influence our private sense of certainty.
The “Partner” Was a Carefully Controlled Illusion
The experiments involved 204 UK adults aged 18 to 50 who completed visual tasks where they judged the direction of moving dots. Participants reported both their choice and their confidence on a sliding scale. Importantly, high confidence earned more points when the answer was correct but lost more points when wrong, discouraging careless overconfidence.
Each experiment consisted of three phases: alone, together, and alone again. In the “together” phase, participants believed they were paired with another person judging the same stimuli. They saw the partner’s choice and confidence, with the joint decision favoring the more confident answer.
However, there was no actual partner. Instead, an algorithm simulated a partner who expressed either generally high or low confidence, while maintaining the same overall accuracy as the real participant. This design isolated the influence of confidence signals, not partner competence.
The first experiment included 80 participants, while the second—a preregistered replication with 124 participants—confirmed the initial findings. Replication is a vital step in establishing the reliability of scientific results, strengthening the study’s credibility.
Confidence Shifts Toward the Partner, Then Persists
During joint decision-making, participants adjusted their confidence ratings toward the algorithm’s displayed style. Those paired with a confident “partner” increased their own confidence, while those paired with an uncertain “partner” decreased theirs. This finding aligns with earlier research demonstrating confidence matching during group decisions.
The more compelling discovery came afterward. When participants returned to deciding alone, their confidence did not revert to baseline. Instead, prior exposure to either a confident or uncertain partner left a lingering effect on their reported certainty. This “contagious” confidence effect, while demonstrated only within the same experimental session, points to how social cues can recalibrate internal confidence benchmarks.
It is important to note the limited scope: the study measured confidence in a specific visual task, and the duration of the effect beyond the experiment remains untested. Whether these shifts in confidence translate into other decision-making domains or persist over days or weeks is still unknown.
Confidence Changed, Not Competence
Higher confidence does not necessarily mean better judgment. The confident and uncertain algorithm partners were equally accurate, and the study found no change in metacognitive sensitivity—the ability to distinguish correct from incorrect choices with appropriate confidence. Participants’ overall certainty shifted, but their actual task performance did not improve or worsen.
This distinction between confidence and competence is critical. Confidence can facilitate action and communication, but without proper calibration, it risks detaching from reality. Someone may sound sure yet be poorly calibrated, while another may sound cautious but accurately gauge what they know.
The study deliberately disentangled these factors by showing that participants absorbed confidence levels from their partner even when those signals carried no additional informational value about accuracy. This underscores how social context can influence not just what we say but how we feel internally about our decisions.
A Cautious Interpretation for Workplace Dynamics
Although the study was not conducted in a workplace environment, its implications resonate with organizational behavior. It cannot claim that a forceful manager will permanently alter team psychology or that colleagues should be chosen based on their confidence levels. However, it challenges the notion that workplace confidence cultures are merely performative.
As highlighted by Silicon Canals in previous discussions about introversion and workplace communication here, rooms often mistake speaking loudly for thinking best. If expressed confidence shifts others’ internal reference points, the dominant confidence style in a room may influence more than just who “wins” a discussion—it may recalibrate everyone’s baseline certainty.
Notably, uncertain partners reduced confidence as consistently as confident partners elevated it. This is not inherently negative. Appropriate caution is valuable when evidence is weak, just as confidence is appropriate when evidence is strong. Problems arise when a group’s preferred expression of confidence becomes disconnected from decision quality.
Insights from the Computational Model
The researchers proposed a computational model where individuals maintain an internal expectation of how confident they should feel. This expectation updates based on discrepancies between anticipated confidence and the confidence reported by themselves and their partner. The model successfully replicated both the rise or fall in confidence during joint decision-making and the persistence of this effect afterward.
It is important to recognize that this model explains the observed behavioral pattern but does not represent a biological mechanism. The study did not employ brain imaging or other neuroscientific methods, and alternative models might equally account for the data.
Interestingly, the effect was consistent regardless of whether participants believed their partner was real, doubted it, or were unsure. This suggests that repeated exposure to confidence signals alone may shift internal confidence benchmarks, even absent a genuine interpersonal connection. However, this might also be a byproduct of the experimental task’s simplicity and repetition.
The Key Unanswered Question: Duration of Confidence Effects
The Birkbeck University report emphasizes that further research is necessary to determine how long socially induced changes in confidence last. Real-world interactions are complex, involving explanations, status dynamics, trust, and feedback—all absent from the experimental design.
Nonetheless, the study’s strength lies in its minimalist approach. By stripping away these complexities, it demonstrated that confidence signals alone can nudge people’s internal sense of certainty. For at least the immediate decisions that followed, some of that borrowed confidence remained.
Understanding the social transmission of confidence has meaningful implications for teamwork, leadership, and decision-making cultures. As organizations seek to foster effective communication and sound judgment, appreciating how confidence spreads and lingers could inform how environments are structured and how feedback is delivered.
For further reading, see the original article Here.
