Whether sharing a weakness is safe depends on the person, the content and the context. The studies examined here did not test safety or whether people underestimate the social cost of vulnerability. They asked how disclosure tracked felt intimacy in everyday interactions and how disclosure related to liking across an older body of research.
The harder and less discussed question is what the sharing actually does. Not whether it is safe, but whether it brings anyone closer, which is usually the reason people are considering it at all. Two studies answer that more precisely than the usual advice does, and the answer has two conditions attached.
Two weeks of conversations, written down immediately
Jean-Philippe Laurenceau, Lisa Feldman Barrett and Paula Pietromonaco asked people to record every social interaction that lasted more than ten minutes, filling in the record straight afterward rather than at the end of the day. The first study followed 69 people for a week, the second 89 people for a fortnight. The results appeared in the Journal of Personality and Social Psychology.
The Artful Age
A weekly letter on aging well, family across generations, and the creative life after the kids leave home.
The headline correlation is strong. How much a person disclosed in a given conversation tracked how intimate that conversation felt, at r = .66 in the first study and r = .55 in the second. How much the other person disclosed tracked it at r = .57 in both. So far this is the familiar picture: open up, feel closer.
The first condition: which kind of disclosure
The second study did something the first did not. It separated what people disclosed into two kinds: facts about themselves, and feelings about themselves.
Disclosing emotion predicted intimacy, with a coefficient of .37. Disclosing facts did not predict it at all, at -.01, which is as close to nothing as these things get. The gap between the two was itself statistically significant.
That distinction is worth holding onto, but only as an association from the second student diary study. The design does not show that adding an emotion statement changes a conversation, and the authors explicitly noted that conversations already experienced as intimate may have prompted more emotional disclosure.
The second condition: what the other person does next
The researchers also measured how responsive each person felt their conversation partner had been: understood, validated, cared for.
Perceived responsiveness partially carried the effect of disclosure on intimacy. In the second study, the link between self-disclosure and intimacy fell from .31 to .22 once responsiveness was taken into account, and the partner-disclosure link fell from .40 to .32.
“Partially mediated” is a statistical description here, not proof of a causal sequence. Disclosure, perceived responsiveness and intimacy were reported by the same participant about the same interaction. Responsiveness accounted for part of their association, but the design cannot show that disclosure caused responsiveness and that responsiveness then caused intimacy.
Three directions, all positive
The older literature adds several directions rather than reversing one. Nancy Collins and Lynn Miller’s meta-analysis in Psychological Bulletin tested three questions: whether people who disclose are liked more, whether people disclose more to those they initially like, and whether people come to like someone more after disclosing to them. All three average effects were positive.
For the first question, across 94 effect sizes from 50 studies, the mean was d = .281. For the second, across 31 effect sizes from 22 studies, the mean was d = .717. The paper also found a positive third effect: people came to like others more after disclosing to them.
The second pooled estimate is numerically more than twice the first, but the two summaries combine different designs, actors and outcomes. They are not a controlled head-to-head estimate of whether disclosure is more consequence than cause. The authors’ conclusion was that the three positive effects operate together in a dynamic interpersonal system.
What these studies cannot tell you
A good deal, and the authors are the ones who say so.
The diary studies are correlational. They cannot establish which came first, and the researchers note that reverse causation cannot be ruled out: perhaps the conversations that already felt close were the ones people disclosed in. They captured one person’s experience of each exchange rather than both sides. And the participants were US college students, with the authors stating plainly that how far the findings extend to other groups is unclear, naming older adults and married couples specifically. That caveat lands hard on a piece written for readers in the second half of life, and it should be read as a limit rather than a formality.
The meta-analysis has its own. Its effects varied enormously by study type, with relationship surveys producing the strongest results and field studies producing negative ones, and its authors could not test properly how the content of a disclosure changed things.
What survives
Not a permission slip. Permission was never the difficulty.
What the evidence adds is narrower. In the second student diary study, emotional disclosure—not factual disclosure—carried the association with felt intimacy, and perceived responsiveness accounted statistically for part of it. That is a useful prompt for conversation, not a rule that facts never matter. The larger literature supports reciprocal links between disclosure and liking rather than one simple direction.