Why a check-in text means more than you think

You think about texting someone, and before your thumb moves you have already run the scenario: it will land oddly, it will read as random, they will wonder why now. So you close the app instead. The hesitation feels like an accurate read of the situation. It is one of the more consistently wrong predictions people make about each other, and there is now a fair amount of research measuring exactly how wrong.

The gap has a name

In 2018, the psychologists Erica Boothby, Gus Cooney, Gillian Sandstrom, and Margaret Clark published a study in Psychological Science documenting something they called the liking gap: after a conversation, people reliably underestimate how much the other person liked them and enjoyed the exchange. It showed up between strangers meeting in a lab, between first-year students getting to know new dorm mates, and between adults meeting for the first time at a workshop. It was not a fragile or occasional effect, and it was larger in people who described themselves as shy — though it did not depend on low self-esteem or a particular fear of rejection. Something more basic seems to be at work: during a conversation you have full access to your own doubts and almost none to the other person's, so you quietly assume the doubt ran in both directions.

The same blind spot, before either of you has said a word

The liking gap describes conversations that already happened. A later line of research asked a harder question: does the same misjudgment apply to the decision to reach out in the first place, before there is any conversation to misjudge? In 2022, Peggy Liu, SoYon Rim, and Kate Min ran a series of preregistered experiments assigning participants to either send an unprompted check-in or receive one, then compared what each side predicted against what the other side actually reported. Across more than a dozen studies, with different kinds of gestures and different relationships, the pattern held: the people initiating contact consistently guessed too low. A message sent just to say hello, or a small unannounced gift, was rated as far less appreciated by the sender than it was rated as appreciated by the person who received it.

Why the misjudgment runs in only one direction

The mechanism the researchers point to is a difference in what each side is attending to. If you are writing the message, your attention is on the message itself — does it sound intrusive, is the timing odd, will it need an explanation attached. You are grading your own performance before you have sent anything. The person receiving it is not grading your performance at all. They are registering something simpler: that on an ordinary day, for no particular reason, someone thought of them. It is specifically the unprompted, mildly surprising quality of it that makes it land, and the research found the effect grew larger exactly when the contact was more out of the blue and between people who were not already talking often — which is precisely the situation that produces the most hesitation before sending.

What this is not saying

It is worth being precise about the limits, because it would be easy to overstate this. The effect is real and it has replicated across many experiments, but it describes an average shift, not a guarantee about any one message to any one person — some of the underlying studies found a smaller effect than the researchers had powered the experiments to detect with full confidence, which is a normal and honest thing for a large replicated finding to disclose rather than hide. None of it means a person who has clearly asked for space should be messaged anyway on the theory that they will secretly be relieved. It simply means that "they probably do not want to hear from me right now" is, on average, a worse prediction than it feels like in the moment of deciding not to send something.

What it means for the message you are not sending

The practical upshot is almost anticlimactic, which is usually a sign a finding is trustworthy rather than oversold. You do not need a reason, a real update, or the right amount of elapsed time to justify reaching out. The version of the message that feels too small or too random to send — the one you type, delete, and decide against — is, by this research, closer to the version that lands best, precisely because it looks like nothing except someone thinking of them with no agenda attached. Most of the friction here is not coming from how the other person would actually react. It is coming from a prediction you are making on their behalf, and that prediction has a well-documented tendency to be wrong in the same direction, for almost everyone, almost every time.