The research on what partners actually prefer is not what you expect

Asking people what they prefer carries a methodological problem that has poisoned this question for as long as anyone has been asking it.

Words make terrible units. Ask whether someone prefers "large" and you've measured their relationship to the word *large* rather than to any physical quantity. Ask them to name a number and you've measured their ability to estimate one, which is dreadful; most people can't accurately estimate the dimensions of objects they handle daily. Show them pictures and you import every framing problem covered in the piece on selection effects.

In 2015 a team including Nicole Prause got around all of it rather cleverly. They 3D-printed 33 physical models across a range of lengths and circumferences, handed them to 75 women, and asked them to choose. Physical objects, in the hand. No words, no numbers, no photographs, no ambiguity about scale.

What they chose

There were two conditions: a one-time partner and a long-term partner.

For a one-time partner, the average selection came out around 16.3 cm in length and 12.7 cm in circumference. For a long-term partner it dropped slightly, to about 16.0 cm and 12.2 cm.

Two things about those numbers, and the second matters more than the first.

They sit above the measured average. The paper says so plainly and I won't soften it, because softening it is exactly the dishonesty this site exists to avoid. The preferred value is meaningfully higher than the pooled clinical mean. If you came here hoping to read that everyone prefers precisely average, that isn't what the study found.

The second point got lost in every write-up I've seen. The gap between the one-time and long-term conditions was small, but consistently in one direction. For an actual relationship the preference moved *down*, toward average. What people pick when imagining a life with someone is lower than what they pick when imagining a night.

That tells you what sort of thing this is. A novelty preference, behaving like the preferences people report for other novel, salient, one-off features: strong in the abstract, weaker as soon as anything real gets attached.

The 52,031-person answer

Now set that 75-person lab study beside the largest survey ever run on the question: Lever, Frederick and Peplau, 2006, 52,031 respondents.

85 percent of women reported satisfaction with their partner's size. 55 percent of men reported satisfaction with their own.

Both studies are true simultaneously, and reconciling them is where the insight lives. In a lab, with no person attached, asked to pick an object, people choose slightly above average. In life, with a person attached, eighty-five percent are satisfied.

No contradiction there. That's what preferences look like once they stop being abstract. Nobody in that 85 percent is reporting that their partner met a specification. They're reporting that the question turned out not to matter much with a whole human being in the room.

The finding I find hardest to write about

Now the uncomfortable finding, uncomfortable in a direction most men don't expect.

Run the numbers back through. 85 percent of women satisfied. 55 percent of men satisfied. 45 percent of men wanting to be larger, 0.2 percent wanting to be smaller.

The distress here isn't distributed the way the folk model says. This isn't partners applying pressure and men absorbing it. It's a thirty-point gap in which men are dramatically more dissatisfied with men's bodies than their partners are.

I want to be careful, because "it's all in your head" is a cruel and useless thing to say to anyone in genuine distress, and that isn't my claim. The distress is real and measurable, and Veale's work showed it tracking depression, anxiety and quality of life.

My claim is narrower and, I think, more useful. The audience you're performing for isn't the audience you think it is. Partners aren't keeping the scoreboard. Other men are, in memory, in adolescence, in a locker room that closed twenty years ago, and so are you, on their behalf, long after every one of them stopped caring.

What the counselling literature says to do about it

There's a body of clinical work on exactly this, and it converges on an approach that's almost insultingly simple.

Show the man the nomogram. Explain the spread. Explain that the figures he has absorbed came from self-report and from selected populations. Then ask what he thinks will change if the number changes, which is usually the question that cracks the thing open, because the answer is hardly ever about the number.

For most men that's enough, and the effect lasts. For a minority with genuine body dysmorphic disorder focused on this feature it isn't enough, and the right treatment is psychological and evidence-based rather than reassurance. Both facts belong in the same breath; leaving out the second is how you make someone feel dismissed.

What the literature never recommends, for a man of normal anatomy, is changing the anatomy.