When someone stops texting you back, your mind tends to construct a fairly detached narrative about what happened on their end. The person on the receiving end of that silence, however, is likely wrestling with something far more uncomfortable than indifference. Research published on 31 August in Personality and Social Psychology Bulletin suggests that those who go quiet experience significantly more negative emotion than those waiting for a reply tend to estimate. Sophia Li, Coral Zheng and Juliana Schroeder conducted three separate experiments examining both sides of message silence, and in two of them, the silent party—whether recalling an actual lapse or imagining one—reported worse feelings than the waiting party had predicted.
What 195 people's actual message histories revealed
The opening investigation recruited 195 adults through Prolific and asked them to examine their own messaging records. Using a between-subjects design, 101 participants recalled a moment when they had stopped responding to someone, while 94 different participants recalled being on the receiving end of such silence. Researchers instructed participants to check their phones rather than rely on memory alone, and to select someone who fell between a stranger and an intimate—neither a close friend nor family member. The median time elapsed since the waiting party's final message was ten days.
The relationships people actually selected were somewhat broader than the instructions specified. When describing the other person, 46.2 per cent called them a friend, 34.4 per cent an acquaintance, 8.7 per cent immediate family or very close friend (a category the instructions had excluded), 4.6 per cent a stranger (also excluded), and the remainder unspecified.
On a scale from 1 ("not bad at all") to 5 ("extremely bad"), those who had stopped replying reported an average emotional toll of 2.27. Those who had been left waiting estimated the other party's feelings at 1.56. This difference, expressed as d = 0.77, was substantial. The waiting group specifically underestimated guilt (2.35 reported versus 1.85 predicted) and sadness (1.84 reported versus 1.46 predicted). They also slightly underestimated relief, at 1.89 reported against 1.56 predicted, which complicates a straightforward narrative of remorse. Of eight emotions measured, these three showed the largest misreading from the waiting side.
The silent party, by contrast, read the waiting side almost precisely: 2.11 predicted against 2.11 reported. Their single significant misreading involved anger, which they slightly overestimated at 1.87 predicted against 1.52 reported—the only emotion running in the opposite direction.
When the silence was unplanned
The gap between the two sides was not uniform across all types of silence. Those who described their non-response as more accidental than deliberate actually felt worse about it, not better; the correlation between intentionality and negative feeling was negative, at r = −.29. The waiting group showed no such pattern. Their predictions remained flat regardless of how deliberate they judged the silence to have been, r = −.002, and the interaction between perspective and intentionality reached statistical significance.
Dividing the sample at the midpoint of the intentionality scale (ranging from "completely accidental" to "completely on purpose") revealed how the misreading shifted. Among the eighty-four participants whose situation fell below the midpoint, those who had gone silent averaged 2.62 while the waiting group predicted 1.50, at d = 1.17. For everyone else, at or above the midpoint, the gap remained but was smaller, at d = 0.52. The researchers emphasise that the misreading was not confined to accidental cases alone.
One unexpected finding contradicted the researchers' own advance prediction. They had anticipated that people left waiting would overestimate how much the silence was a response to something they themselves had done. Instead, in these recalled cases, the direction reversed: those who had gone quiet attributed more intentionality to their own behaviour (3.75) than they were credited with (3.19).
Testing a simple intervention
The second experiment placed 601 people into a written scenario rather than asking them to recall their own experiences, and introduced a practical comparison. Each participant received one side of an exchange and viewed two versions: one where the friend simply did not reply, and one where a single line came back: "Sorry, but I can't talk right now". The perspective effect reappeared, accounting for roughly fifteen per cent of the variance. The one-line reply did reduce the misreading. When imagining no response at all, participants put their own bad feeling at 3.23 while predicting the other side would estimate 1.94. When the brief reply appeared, they put it at 2.33 against a prediction of 1.82. The gap shrank but remained statistically significant at d = 0.49. The short message narrowed the distance between felt and perceived emotion without closing it entirely.
The brief reply also shifted how the waiting participants explained the silence. When nobody replied in the scenario, they split almost evenly among three supplied reasons: no bandwidth, the conversation wasn't interesting, or they found the waiting person annoying. When the one-line reply arrived, 57.7 per cent attributed it to lack of bandwidth and 19.7 per cent to being annoying. Seven words of acknowledgement moved the waiting participant's explanation away from self-blame.
The live conversation study
For the third experiment, the researchers constructed the situation rather than describing it. Six hundred people were recruited. Fifty never matched or encountered technical issues, and another ninety-two were lost when they or their partner timed out mid-study, leaving 458 people in 229 complete pairs. Each pair had conducted an eight-minute get-to-know-you conversation using questions designed to build closeness rapidly. One partner then faced a choice: participate in a second five-minute conversation where they would discuss their own personality and write a brief summary, or skip it and proceed to the study's end for the same payment. The researchers positioned skipping as the more appealing option, since only pairs where someone left would enter the main analysis. In 175 pairs someone took it, and those 350 people formed the sample for primary analysis.
An important note: the scale used in the third study differed from the first two. The first two ran from "not bad at all" to "extremely bad", while the third ran from "extremely good" to "extremely bad", making direct numerical comparison impossible. Several other measures in this study used their own scales.
The headline finding showed no significant gap. People left sitting in the chat window predicted their partner felt 3.35 bad about it; the partners reported 3.26. The paired difference did not reach significance, at p = .357, and the confidence interval still allowed for a small gap in either direction.
The gap reappeared in two other places within the same study. One was a preregistered secondary analysis of the 54 pairs where the potential leaver stayed and completed the personality interview. Asked how they would have felt had they left, they put it at 4.22 against their partner's prediction of 3.24, at d = 0.93. The other emerged in a subgroup of pairs where someone did leave, divided by how much each person considered the behaviour to be ghosting—a moderator the authors acknowledge they had not registered in advance. Among the 121 participants who rated it above the midpoint (46 who left and 75 who were left), the underestimation returned at 3.93 against 3.51, at d = 0.43. These were not paired figures, and the paper provides no comparison for those below the midpoint. A more inclusive cut including those exactly at the midpoint produced 3.72 against 3.43, at d = 0.30. The leavers' bad feelings climbed with how strongly they read their own silence as ghosting, r = .42, while the waiting side's predictions stayed nearly flat against their own reading of it, r = .13.
What did survive the live test, and survived it substantially, was a different misreading. On a scale from "extremely positive" to "extremely negative", people left waiting put their partner's feelings towards them at 3.41; the partners reported 2.29, at d = 0.91. On a scale from "not at all" to "very much", the waiting side put their partner's interest in friendship at 1.53, against a reported 2.58. The person who left the chat window was, on average, warmer towards the person left in it than that person had guessed.
How the two sides named what happened
The live study produced something the first two could not, because both halves of the same interaction were on record. They disagreed about what had occurred. Asked how much the behaviour counted as ghosting, those who left averaged 2.56; those they left averaged 3.06. The leavers had been reminded repeatedly that they had chosen to skip the conversation and leave someone waiting alone without explanation. They still scored it lower.
The study's boundaries
The authors are explicit about their limitations. A short online study, they note, cannot fully capture the stakes of a long relationship deteriorating. Their participants had known each other for eight minutes and, the paper suggests, most likely expected never to meet again. The waiting partner had been told in advance that the other might not return, which may have reduced the uncertainty typically accompanying ghosting. Together, these factors made the study a relatively conservative test of their hypothesis. Separately, in the limitations section, the authors acknowledge that assigning ghosting to people in an experiment raises both ethical concerns and internal validity problems. They propose embedding short surveys in messaging platforms as one way to study the real phenomenon.
What the research ultimately rests on is five-point scales completed by people recalling a text thread, people imagining one, and people who had met eight minutes earlier, with materials, data and analysis scripts posted openly. This suffices for claims about averages, with the paper's own scope attached: in recalled experiences, in imagined scenarios, and in the live conversation where the non-response was recognised as ghosting, the state behind a non-response is more uncomfortable and less indifferent than the person waiting assumes. In the scenario study, a short acknowledgement narrowed the distance between the two readings. Why the person you are thinking of has not replied remains outside the study's reach. The samples consisted of US adults describing acquaintances and friends they were not close to, plus strangers paired for eight minutes—not your friend. And if repeated silence from someone close to you is causing real harm, a therapist or counsellor is the appropriate resource.
What you can actually use
The practical takeaway fits in a sentence. If you are the one who has gone quiet, a one-line acknowledgement was the element that closed part of the gap in the scenario study, though that study never varied what went into the line: the apology, the reason and the implied later all arrived together, and the authors say they cannot determine which part did the work. If you are the one waiting, the flat, indifferent version of the other person you have constructed is probably a worse likeness than the data supports; misreading someone as indifferent could, the authors argue, reduce the chance of anyone picking the thread back up.
What would settle the question is a comparison the paper does not report: both members of a pair rating the same silence as ghosting, and their feelings weighed against each other within that pair. The authors' own proposal for stronger evidence points away from the lab entirely, towards short surveys embedded into the messaging apps where the silence actually happens.
Source: Silicon Canals
