<link rel="stylesheet" id="wp-block-library-css" href="https://c0.wp.com/c/5.8.2/wp-includes/css/dist/block-library/style.min.css" type="text/css" media="all">
loader image

I recently spied an article in my Google Scholar alert feed that had a rather interesting abstract.

This study investigates spousal correlation and mate preference for height in the Japanese population…It also indicates a possible role of a sexual imprinting-like mechanism in human mate choice.

The research found that Japanese university students’ ideal marriage partners were of a similar height to their parents. This, they suggest, indicates people may “imprint” on their opposite sex parent, seeking out partners of a similar height to them.

I was immediately skeptical of the whole thing. It seems paradoxical given that all other research indicates we prefer people with a different genetic make-up; so why would we imprint on and seek out people similar to our parents, who obviously have very similar genetics to us.

So, armed with caution I read on.


A series of questionnaire surveys were conducted on [616] Japanese students…[asking] the height of the subject, the heights of his/her parents and either the ideal partner’s height.

It’s your fairly typical psychology method: professor asks his students questions because they’re the group he has a lot of access too. Here a few potential issues arise, but most can be dismissed.

Firstly there’s the fact that these are university students being questioned. Can they be extrapolated from this subset to everyone else? Being such a student, I can attest to the fact we are a wildley different breed, with unique fur, sleeping patterns and diets. But alas, we are not so different that the results can’t be attributed to society at large.

Secondly, there’s the fact that the sample is unevenly balanced, with more women being questioned than men. However, the paper examines each gender individually, so this becomes a non-issue.

Thirdly, there’s the fact this relies upon self-report rather than actual heights. The paper notes a prevalence of people saying they, their parents or ideal spouses have a height ending in an 0 or a 5. This is probably because they can’t remember the exact height, which introduces a potential error into the equation.

Fourthly, they offer no way to control for the economic background of the parent, which other research suggests does impact the kind of data they are gathering here.

There’s also the fact the sample size is quite small, which can only be dismissed by further supporting data. So whilst the methodology is good for an initial investigation, these results can only really be taken as preliminary until there is more support. Preferably support that includes exact heights rather than remembered ones and controls for more variables.


the height difference between the subject and his/her opposite-sex parent positively explained the ideal height difference, especially for males

Basically, the height difference between someone and their opposite sex parent was similar to the height difference between them and their ideal partner. If a woman’s father was shorter than them, they wanted a shorter husband. If he was taller, they wanted a taller husband.

They also found that males also preferred partners with an absolute height similar to their mother. If she was taller than them, not only would they like a taller wife but a taller wife who is of a height similar to the mother. Females showed a similar relationship with their father, although it was much weaker.

The statistics behind all these results appear, at least to a non-statistician like myself, to be valid if not as strong as one would hope. Indeed, if the earlier issues with the methodology mentioned alter some of the statistical results by as little as 2%, they would be considered “weak.”

Now, a weak statistical relationship is still a relationship, so the conclusion does seem sound. Unless methodological issues to drastically overhaul the data, it would seem the researchers are right.


So, does a mild/weak statistical relationship between the heigh difference of an idealised partner and a subject and the height difference of an opposite sex parent and a subject mean there is imprinting?

The phrase I think you’re looking for is “not really.”

Whilst this research does establish there is such a relationship, it doesn’t do any investigation into the mechanism behind it. So pretty much any explanation for this observation is as valid as any other. Imprinting is just one of many potential explanations, none of which are confirmed or denied by this research.

The thing is, the article seems to know this. In the conclusion it discusses a variety of potential explanations, suggesting why each might be plausible. Imprinting is just slipped in at the end, in a kind of “here’s some science…btw, imprinting” way. And yet that’s what gets put in the abstract to lure suckers like me into reading it.

Seki, M., Ihara, Y., & Aoki, K. (2012). Homogamy and imprinting-like effect on mate choice preference for body height in the current Japanese population Annals of Human Biology, 39 (1), 28-35 DOI: 10.3109/03014460.2011.635695

Related posts


pithecanthropus sapiens · 29th July 2017 at 7:24 pm

I believe there are similar imprinting-like effects with race and age difference. People prefering/tending to marry partners of the same race of the opposite-sex parent, and with a similar age gap/proximity that their parents had, also following gender.

    Adam Benton · 1st August 2017 at 2:54 pm

    I’ve heard that’s the case, but never really looked into it that much. Are there any good sources you know of?

Leave your filthy monkey comments here.

This site uses Akismet to reduce spam. Learn how your comment data is processed.