INTERPERSONAL RELATIONS AND GROUP PROCESSES

58

firm previous reports, cited earlier, that spousal pairs tend to
resemble each other weakly to moderately on, in this case, 88 of
88 mainly psychological variables. These 88 variables are not
mutually orthogonal, however; the recreational and vocational
interest scales correlate with one another and with the personal-
ity, talent, and self-appraisal factors, the attitude items, and
with years of education. As others have reported, the strongest
spousal similarities are in the related dimensions of tradition-
alism and religiosity, followed by educational attainment and
stature. The question that remains to be addressed is this: Does
this positive homogamy, or vzwi∕απ⅛ model satisfactorily ac-
count for specific mate selection?

Consider an unmarried, 30-year old man hoping to select a
mate from among all the eligible women of his acquaintance
between the ages of (say) 20 and 35. If he has special ethnic or

religious requirements-if he would consider only Norwegian
Lutherans, for example-then he will have further narrowed
what K∣rckhoff (1974) would call his “field of eligibles.” As-
sume we ‘know that the spousal correlations for physical attrac-
tiveness and traditional values are both about .50 and that these
variables are orthogonal and normally distributed; how will
this information help our seeker focus his search? Suppose his
own scores on both attractiveness and traditionalism are ex-
actly at the mean (which we will set at 50 ±10). We know
that the standard deviation of his potential mates on both vari-
ables will be 10(1 — r²)^ιz2 = 8.7. If all seekers take care to select
mates who differ from themselves in traditionalism by not
more than 2.5 (8.7) = 21.75 T-score units, then the spousal
correlation for traditionalism will remain at about .50 and we
know that about 99% of all potential mates for our seeker will