Where to Sit in a Meeting, According to the Actual Research

Search for advice on where to sit in a meeting and you will find the same things repeated: the head of the table is the power seat, sit to the right of the boss, avoid the corner. Most of that comes from nowhere. A small part comes from real research conducted between 1950 and now, and that part is almost never cited.

This article separates the two.

What holds up

The person opposite you is the person who answers you.

In 1950, Bernard Steinzor published in the Journal of Abnormal and Social Psychology the observation that in small groups seated in a circle, the greater the distance between two people, the greater the chance one speaks immediately after the other.¹ You respond to the person across from you, not the one beside you.

Worth reproducing the author's own caution: he wrote that the evidence confirmed the hypothesis "in part." It is not a law.

The application is direct. If there is someone whose opinion you need to surface, do not sit next to them — sit opposite. And if you want two people to talk to each other rather than to you, place them across from one another.

The head of the table has a real effect.

In 1961, Fred Strodtbeck and Harmon Hook published in Sociometry a study of twelve-person juries at a rectangular table. Those seated at the ends were chosen as foreman disproportionately often, and people at the ends and the middle participated more than those along the sides.²

One note of honesty: the exact figures from that study sit behind paid access and we have not read them at source. The finding is solid and has been cited for sixty years; the specific percentage we will not publish second-hand.

Table shape changes which argument works.

This is the most interesting finding and the least known. Juliet Zhu and Jennifer Argo published three experiments in the Journal of Consumer Research in 2013 showing that seating geometry activates different needs:³

• in a circular arrangement, people evaluate collective appeals more favourably — family-oriented cues, majority endorsement;
• in an angular arrangement, they evaluate individuality appeals more favourably — self-oriented cues, minority endorsement.

The proposed mechanism is that circles prime a need to belong and angles prime a need to be unique.

If your pitch rests on "everyone in your industry already does this," you want a round table. If it rests on "this will set you apart," you want an angular one.

What does not hold up

Power posing. This is the most instructive case in recent applied psychology, and it is still repeated in corporate training as fact.

The original study — Carney, Cuddy and Yap, 2010, Psychological Science, with 42 participants — claimed that two minutes of expansive posture raised testosterone, lowered cortisol and increased risk tolerance.⁴ It became one of the most-watched talks in history.

In 2015, Ranehill and colleagues replicated it with a sample roughly five times larger. The hormonal effect did not appear. The behavioural effect did not appear. Only the subjective feeling of power replicated.⁵

In 2016, Dana Carney — first author of the original — published a statement saying she no longer believes the effect is real, describing the methodological choices that produced it. She told Retraction Watch that the data were real, the effect does not replicate, and that she updated her beliefs as the evidence accumulated against it.⁶

In 2018, Cuddy, Schultz and Fosse analysed a larger body of work and argued that the felt-power component holds up while the hormonal and behavioural components do not.⁷

What survives: an expansive posture may make you feel more confident. It does not change your hormones, it does not improve your performance, and it is not read by others the way it was promised.

Interpersonal distance as an exact number. Edward Hall's work on proxemics is an influential and useful frame, but the specific boundaries vary considerably across cultures and are not robustly replicated. Use it as a concept, not a measurement.

And, inevitably, the 93% figure. Much of the seating and posture content anchors on it. We dealt with that in The 93% Myth.

And when the meeting is on video

Almost everything above ceases to exist. There is no head of a grid of equal rectangles, no physical distance, and the Steinzor effect has no geometry to operate in. What remains is a different set of channels, covered in Reading the Room When There Is No Room.

Worth noting the implication: in-person meetings became less frequent and, precisely because of that, more decisive. The few physical meetings left now concentrate decisions that used to be spread across several conversations.

The rule that summarises it

There is little proven about positioning in meetings, and it is less glamorous than the folklore: sit opposite the person you need to hear, understand that the ends of the table concentrate attention, and choose the table shape to match the kind of argument you are about to make.

The rest — the power chair, the hand placement, the two minutes of posture before walking in — is performance. Performance has a real cost: it consumes the attention you would need in order to listen.