When I am about to meet someone new, I often prepare for the wrong job. I think about what I might say, which story might land and how to avoid an empty pause. It is a very ordinary form of social performance. Yet the conversations I remember warmly are often the ones in which somebody noticed a small detail in my answer and stayed with it.

A 2017 paper in the Journal of Personality and Social Psychology, led by Karen Huang at Harvard, offers one explanation. Across controlled chats and a speed-dating dataset, people who asked more questions, especially follow-up questions, tended to be liked more by the person answering them. This is one paper, not settled consensus. Its controlled experiments and its field evidence also deserve different levels of confidence.

The experiments changed one small part of the conversation

In the first experiment, 398 people who completed the study were paired with strangers for 15-minute instant-message conversations. One person in each pair was privately instructed to ask either at least nine questions or no more than four. Their partner knew nothing about this instruction.

The partners later rated how much they liked the person they had spoken with. On a seven-point scale, people paired with the high-question askers gave an average liking score of 5.79, compared with 5.31 for those paired with low-question askers. That difference mattered within the experiment, but it was not enormous. It does not mean a question can rescue a conversation with no warmth, curiosity or shared ground.

A second experiment analysed 338 people in 169 pairs. This time, both partners were independently assigned to ask many or few questions. Once again, people liked partners who had been assigned to ask more questions. Because the researchers randomly changed question-asking, these experiments can support a causal claim within the rather specific world of short, typed conversations between strangers.

The follow-up mattered more than the question mark

The researchers then had six assistants classify the questions in 368 transcripts from the first two experiments. They separated follow-ups, which stayed with something the other person had just said, from full-switch questions that moved to an unrelated subject, along with several other types.

Follow-up questions were the only category that predicted greater liking. People told to ask many questions also devoted a larger share of those questions to follow-ups: about 41.5 per cent, compared with 28 per cent among those told to ask few.

That distinction is more useful than the simple instruction to “ask questions”. “Where did you go next?” works differently from abruptly asking, “What do you do for work?” The first depends on the answer that came before it. It shows that the conversation has a memory.

“Feeling heard” has a narrower meaning here

The paper uses the term perceived responsiveness. In this literature, that means the extent to which someone seems to listen, understand, validate and care. The high-question askers were rated as more responsive, and the researchers’ statistical mediation analysis suggested that this perception helped explain why they were liked more.

That is the evidence beneath the phrase “made the other person feel heard” in the title. It is a fair plain-language summary, but the study did not peer inside anyone’s mind and prove a single mechanism. Mediation is evidence consistent with an explanation, not a camera recording the explanation as it happens.

An independent 2014 experiment on active listening points in a similar direction. In conversations involving 115 undergraduates and trained partners, people who received active-listening responses felt more understood than those who received advice or simple acknowledgements. They also reported more satisfaction and social attraction than people who received only acknowledgements. The effects were small, however, and active listening was not superior to advice on every measure.

The observer saw a different conversation

One of the best details in Huang’s paper is that outsiders did not respond in the same way as the people inside the exchange. Separate observers read transcripts and did not especially prefer the high-question askers. If anything, the answerers could seem more interesting because they supplied more of the visible content.

The benefit belonged to the person being asked. A good follow-up may not make the asker look dazzling from across the room. It changes the private experience of the conversation by showing that an answer has been received rather than merely waited out.

This sits neatly beside something I have written before about small talk. The opening question does not always need to carry much information. Its value may be that it creates a place for the next, more attentive question.

The speed-dating result is interesting, but contested

The paper’s third study used recordings from 110 people at heterosexual speed-dating events. Each person went on 15 to 19 four-minute dates. The analysis covered 1,961 date observations, matching the transcripts with each partner’s decision about whether they wanted another date.

A computer model estimated which questions were follow-ups. People with a higher follow-up-question rate received more second-date offers. The original estimate suggested that an eight-percentage-point difference in follow-up-question rate was associated with one additional successful match across 20 dates.

Associated is the key word. Nobody was assigned a question style on these dates. Chemistry, attraction and the particular pairing could influence both the questions asked and the decision made. In 2019, Avraham Kluger and Thomas Malloy published a reanalysis challenging the speed-dating conclusion. Using a model designed to separate the person, partner and particular pairing, they reached a different result and argued that a second-date offer is not identical to liking. The original researchers replied that their central conclusions still held. That disagreement does not erase the randomised chat experiments, but it makes the dating claim something to hold more lightly.

A conversation is not an interview

There is an obvious limit here. Questions can be intrusive, mechanical or exhausting. The original paper itself suggests that the relationship may not continue in a straight line forever. Someone who asks endlessly while revealing nothing may leave the other person feeling examined rather than known.

The grounded reading is not that the most likeable person counts question marks. It is that people can become so busy preparing their next contribution that they fail to build on the contribution already offered. A follow-up question is useful precisely because it cannot be fully prepared in advance.

I still enjoy a good story, and I do not think the paper asks us to become blank conversational mirrors. It simply shifts some of the weight from performance to attention. The person we enjoy talking to may not be the one who brought the best material. It may be the one who made our own material feel worth continuing.