Evidence status: conditional support Experiments, longer-term observations and team studies find that after getting acquainted, people tend on average to underestimate how much their partners liked them and enjoyed the conversation. This “liking gap” is a group tendency, not evidence that any particular person likes you, and it cannot override explicit rejection, boundaries or power differences. This is a version 0.1 research draft, with evidence reviewed to August 2026.
How Everyday Psychology Takes Shape · Season Two: How Do We Read Other People? · Article 6
After a first conversation, the details most available may not be the other person’s nods and smiles. We remember our unnecessary explanation, several seconds of silence and a sentence that could have been phrased better. We liked the other person, but are less sure that the feeling was returned.
Both people can leave with this impression. Each knows their own positive response and thinks that it was not fully reciprocated. There is no logical contradiction because each is estimating a different object: their position in the other person’s mind.
In 2018, Erica Boothby, Gus Cooney, Gillian Sandstrom and Margaret Clark named this average tendency the liking gap. The finding is not that everybody is liked. Across several getting-acquainted settings, participants reported liking their conversation partner more than they estimated that the partner liked them. Because the two people were paired, each person’s actual report supplied a comparison for the other person’s estimate.
From laboratory conversations to developing relationships
The researchers asked strangers to have short, medium and longer conversations, then report their liking, enjoyment and beliefs about the partner’s evaluation. A gap appeared across conversation lengths. Independent observers could sometimes identify the positive signals that partners expressed on video, suggesting that liking was not entirely hidden. People inside the interaction were also occupied with monitoring their own performance.
The paper followed first-year university students living with new roommates. The effect was not confined to a five-minute stranger task. It could persist during an early phase of relationship development and then diminish as roommates accumulated clearer evidence. A sample of adults meeting in a UK personal-development workshop showed the same general direction.
In 2021, Adam Mastroianni and colleagues extended the work to small-group conversations, student engineering teams and workplace teams. Members tended to underestimate how much other members liked them. The gap was associated with self-reported communication, team efficacy and job satisfaction. These studies broaden the settings, while much of the literature still comes from a limited set of research groups. Cultural and age generalisability need continued testing.
A developmental study found a similar gap emerging among children from around age five, approximately when concern about other people’s evaluation becomes more prominent. That is an informative group pattern, not a psychological switch that activates on a child’s birthday.
Why our own negative details sound louder
Conversation partners possess asymmetric information. You hear your internal editor: Did I speak too much? Was that question dull? The other person receives only what became observable. At the same time, you directly know that you liked your partner and must infer their liking from courtesy, expression and response.
Defensive pessimism may contribute. Lowering an expectation under uncertainty can seem to soften possible rejection. That idea cannot by itself carry the whole effect. One recurring account in the studies is that people focus heavily on imperfections in the impression they made and underweight positive cues in their partner’s experience.
Social norms make signals ambiguous. A smile can be courtesy or genuine interest. A short answer can reflect time pressure rather than dislike. When neither person wants to appear over-eager, real liking is not made fully public, and uncertainty can persist.
The result belongs to the same broader problem seen elsewhere in this season: a person has rich access to self-critical internal information and only partial access to the other side. More thinking does not necessarily add more interpersonal data.
An average optimistic result cannot read a specific mind
“People like you more than you think” is an attractive reassurance and an easy overstatement. The evidence is an average difference in samples. In a particular interaction, a partner may genuinely lack interest or explicitly reject further contact. No psychological effect can cancel “please do not contact me”, expressions of discomfort or sustained avoidance.
Power changes signals. A subordinate’s friendliness to a supervisor may partly reflect the role. A service worker’s smile is part of paid labour. Speed-dating interest and collegial cooperation have different criteria. Reading occupational courtesy as private attraction is not an application of liking-gap research.
On the other hand, an absence of exuberance does not automatically mean rejection. People can be anxious, tired, speaking a second language or following different cultural norms around gaze and praise. If continuation matters, a respectful, low-pressure action provides better evidence than indefinite analysis of expression.
One light test is more informative than repeated replay
If a conversation felt worthwhile, a specific message with low demand can create new information: refer to a shared topic and offer a next step that can naturally be accepted or declined. The response contains more evidence than another solitary replay of the silence.
Teams can make positive feedback explicit. “Thank you for clarifying that issue” works better than assuming a colleague knows they are valued. The organisational implication is not that managers should require affection. It is that cooperation should not depend entirely on members guessing whether they are accepted.
We can also inspect the evidence used in post-conversation review. Does a negative detail come from an observable act or from “I felt awkward”? What did the other person actually do? Is there a direct next step capable of testing the inference? Separating internal experience from interaction evidence is more honest than repeating “they definitely liked me”.
The same method protects against the opposite error. If the other person repeatedly declines without proposing alternatives, a theory about average underestimation should not be used to preserve a preferred conclusion.
The liking gap explains some systematic pessimism; it does not convert ambiguity into consent or make one friendly conversation carry a relational commitment. A low-cost test is defensible only when a negative answer is allowed to change the judgment.
How mutual uncertainty can prevent a relationship from forming
The liking gap matters for more than reassurance. If both partners underestimate how much they are liked, each may send one fewer invitation or ask one fewer question. The relationship is not explicitly rejected. It fails to continue while both wait for confirmation. A mistaken metaperception changes the next action and helps create the later fact that “we were never close”.
A relationship therefore does not exist only in two private states of liking. Liking must enter a shared structure through intelligible signals before it can support another interaction. When a feeling is not expressed, or its expression is not recognised, a private state struggles to become a relational fact.
The version 0.1 conclusion is that modest scepticism towards pessimistic post-conversation review is evidence-based. A partner may have enjoyed the exchange more than you estimate, especially when the estimate is built from your own small mistakes. The research grants nobody a licence to assert that another person secretly likes them. The justified application is to lower the threshold for one low-cost, boundary-respecting follow-up and let an actual response replace mind-reading. If the response is a clear no, accept it.
Primary research sources
- Boothby, E. J., Cooney, G., Sandstrom, G. M., & Clark, M. S. (2018), “The Liking Gap in Conversations”.
- Mastroianni, A. M., Cooney, G., Boothby, E. J., & Reece, A. G. (2021), “The Liking Gap in Groups and Teams”.
- Wolf, W., Nafe, A., & Tomasello, M. (2021), “The Development of the Liking Gap”.
- Tissera, H., Gazzard Kerr, L., Carlson, E. N., & Human, L. J. (2021), “Social Anxiety and Liking: Towards Understanding the Role of Metaperceptions in First Impressions”.
Series navigation: How Everyday Psychology Takes Shape — Season Two overview
Discover more from Geoffrey Chen
Subscribe to get the latest posts sent to your email.