Assortative marriage means the tendency for people to choose
spouses who resemble each other more closely than one would expect by
chance. It may also promote social diversity, and govern the
"inheritance" of social traits such as religion and voting patterns which
are not genetic but are strongly familial.
A study of assortative marriage was undertaken in a series of Edinburgh couples, in order to determine
a) the strength of assortative marriage for a wide range of physical,
psychological and social traits.
b) to what extent these similarities were present at the time of
marriage, and whether the couples had converged or diverged since
c) the nature of the interaction of those traits which were interlinked.
d) the implications of these similarities for those genetic and social
traits which assortative marriage is believed to influence.
By confining the study to those traits which can be quickly and ob¬
jectively measured, it was possible to study a wide variety of traits and
yet have a large sample size. Eventually 68 engaged couples, 113 newly
weds, and 222 couples who had been married for five years or more
were ascertained and visited at home.
As many as possible of these 403 couples were then revisited a year
later, to remeasure the traits. A computer program was written to
handle the large volume of data collected.
General traits measured included age at marriage, fertility of the
couple, number of sibs of each partner, and their previous marital status.
Age was highly correlated, as with other studies. The correlation was
higher in remarriages than first marriages, and lower in old than young
Fertility, i.e. total planned family size was of course strongly cor¬
related; there was also evidence of a marked reduction in plans for total
planned family size as the marriage progressed. Family plans were not
related to education or social class.
There was little or no association for number of sibs. Previous
marital status was highly correlated even after correction for age. In
young couples, bachelor with spinster and divorcee with divorcee were
preferred combinations, and widows/widowers showed no preference. This
was probably related to previous fertility. In older couples, widow with
widower was preferred but there was no distinction between singletons
Physical traits measured included height, weight, skinfold thickness,
blood pressure, and pulse before and after exercise. Height was cor¬
related even after taking account of age and this leads to an increase in
the population variance for height. However there is a complex
interaction with perceptions of social class. Weight, assessed in a number
of ways, did not show association in this sample. The various cardiovascular
traits occasionally showed correlations but these all proved to be
artefacts of age, smoking etc.
Psychological traits studied were neuroticism and extroversion, as seen
in oneself and in one's partner and measured by Eysenck's Personality
Inventory. Couples showed assortative marriage for neuroticism; curiously
they assorted for extroversion only in newly weds. It was expected that these
psychological traits might show convergence during the course of the
marriage. In fact there was no evidence of this; the results for extroversion
may even have indicated divergence.
Social traits examined were occupation, father's occupation, smoking
habits, education, and religious practice. There was assortative marriage for
all these traits but all were highly intercorrelated. Initially occupation
appeared to be an independent factor but during the course of the survey it
behaved more and more as if determined by education. The explanation
seemed to be that in recently married couples there was a delay of some
months before they found work to match their educational achievements. The
other traits were assorting independently. One might have expected, say,
smoking and religion to show some convergence but no such tendency was
evident in the overall sample. These social traits all play a part in selecting
a partner therefore, but do not tend to converge or diverge markedly there¬
The main correlations found were in age, height, neuroticism, education,
smoking habits, and religious practice. Assortative marriage for these traits
may well play a role in cultural adaptation to change.