The news in brief
Three studies by researchers at Berkeley and Cambridge (more than 1,100 participants) show that people who are ghosted underestimate how uncomfortable the other person feels, especially when the silence was not intended. A very short reply such as “sorry, can’t talk” is enough to reduce the misunderstanding.
What the study says
Sophia Li and Juliana Schroeder, of the Haas School of Business at Berkeley, and Coral Zheng, of the Judge Business School at Cambridge, published online on 31 August 2026 three experiments in the journal Personality and Social Psychology Bulletin (Li et al., 2026). Their question: when someone stops replying, does the person waiting guess what the silent one feels?
We read the article in full, in its accepted version made open access by the University of California. The researchers adopt a broad definition: an unexplained silence that at least one of the two thinks could be permanent, forgetfulness and temporary silences included.
The three experiments
The first, conducted in 2023, brought together 195 adults recruited online, randomly assigned to recount either the last time they had not replied to someone (101 people) or the last time someone had not replied to them (94 people) (Li et al., 2026). These were acquaintances or friends, neither strangers nor close intimates. On a scale from 1 to 5, those who had gone silent said they felt bad at 2.27, whereas the ghosted imagined them at 1.56 (Li et al., 2026). The ghosted underestimated their guilt, their sadness, and also their relief. The gap grew when the silence was more accidental: 2.62 against 1.50 (Li et al., 2026).
The second, conducted in 2025, had 601 adults imagine an exchange with a friend they did not yet know well (Li et al., 2026). Each read two versions: the friend does not reply at all, or writes only “sorry, I can't talk right now”. Faced with silence, the one who went quiet imagined feeling bad at 3.23, and the one waiting believed it was 1.94; with the brief note, the gap narrowed to 2.33 against 1.82 (Li et al., 2026). The brief note also changed the interpretation: 57.7% of those who received it thought the other simply lacked time, against 34.3% of those who received nothing (Li et al., 2026). According to the authors, the one waiting then blames themselves less for the silence.
The third, conducted in 2024, paired 229 pairs of strangers for a written conversation of eight minutes (Li et al., 2026). One of them could then skip the second conversation without warning, while the other waited five minutes. The analysis covers the 175 pairs in which there was silence, or 350 people (Li et al., 2026). To the authors' surprise, this time there was no overall underestimation, 3.26 against 3.35, a non-significant difference (Li et al., 2026). On the other hand, those who waited believed the other was far more hostile towards them than they were, 3.41 against 2.29 (Li et al., 2026). When both truly judged the behaviour to be ghosting, the underestimation came back: 3.93 against 3.51 (Li et al., 2026).
The figures to remember
The three studies total 1,146 people analysed (Li et al., 2026). The scale of the third study ran from “extremely well” to “extremely badly”, which rules out comparing its averages with those of the other two. In the third study, the one who had slipped away labelled their behaviour as ghosting less than the one who waited, 2.56 against 3.06 out of 5 (Li et al., 2026). And the more they saw it that way, the worse they felt.
How to read these results
The authors acknowledge several limits. All participants lived in the United States, and cultures differ in their taste for frankness or for indirectness (Li et al., 2026). The situations studied mostly concern the early stages of relationships and brief exchanges, not the end of a long love story. In the live experiment, each knew they would probably never see the other again, and the one waiting knew their partner might not come back. The role of the perception of ghosting, not planned in advance, needs confirmation.
The data are mostly self-reported, memories or imagined reactions. And these are averages: they do not say that the person who left you without news necessarily feels guilty. They may feel none at all.
Finally, these experiments show a gap in perception, not a moral rule: they do not excuse disappearing on the grounds that the one who leaves suffers too.
What this changes for the reader
For the one who waits, the study offers some perspective, not a promise. Knowing that silence is often accompanied by discomfort in the other, and that, in the live experiment, they judged the other less harshly than we believed, can help you not read it as a verdict on yourself. But your peace of mind does not have to depend on their regrets.
For the one who no longer replies, Coral Zheng advises, in PsyPost, sending a short reply before going silent. Which elements of this note matter most, an apology, an explanation or a promise to come back, remains to be studied, the authors say. Nothing obliges you to reply when you feel unsafe or under pressure: cutting off can then protect you.
Two people who no longer speak often live very different silences. One imagines indifference, the other sometimes struggles with discomfort. A short note does not settle everything, but it can keep each person from inventing the answer alone.
What our main page adds
Our main page completes this finding with an overview of the practice. It distinguishes the slow fade, where replies become further apart until they vanish, breadcrumbing, which keeps interest alive with a few messages without offering anything, and orbiting, where someone has disappeared from the conversations while still watching your posts. It explains why people disappear: to avoid an unpleasant refusal, to invest little after a few messages, to run several conversations at once, or out of fear after an insistent message, a case where cutting off is the right response. To end a relationship yourself, it suggests a message of about thirty seconds in three parts, a thank-you, a clear sentence and no door left ajar. Lastly, it recalls that you can cut off without explanation in the face of insults, threats or a request for money. All of this is in the page on ghosting.
Going further
This news item is part of a broader topic, which our page Ghosting: disappearing without a word covers in detail.