Susan L. Trumbetta1 & Kay Phillips1
Resolution of the origins of observed marital correlations is required in order to assess the influence of assortment on both familial resemblance and within-person correlations in offspring. Assortment can also affect the distribution of genes and environmental variables in the population. When assessing assortative mating for correlated traits, examination of marital correlations alone can be uninformative. Delta path methods take into account trait correlations within persons and estimate the extent to which mates assort directly on specific phenotypes (Van Eerdewegh, 1982, Statistical Selection in Multivariate Systems with Applications in Quantitative Genetics, Washington University, St. Louis, Missouri, unpublished doctoral dissertation; Carey, 1986, AmJHumGen, 39, 775-786). We examined assortative mating in a sample of 148 parent couples participating in the longitudinal Louisville Twin Family Study. Data from nine scales of the MMPI-2 were analyzed using maximum-likelihood estimation in a nested series of multivariate delta path models. The results indicated significant direct isomorphic assortment on traits measured by four of the scales: Scales 4 (psychopathic deviance), 2 (depression), 9(mania), and 0 (social introversion). Direct cross-trait assortment was not significant, and may require increased sample size to be detectable.
Address: CDU, MDR-111, HSC, Louisville, KY 40292-0001
1Department of Pediatrics, University of Louisville Medical School, Louisville, KY 40292 Phone: 502-852-3278 Fax: 502-852-3277 E-mail:sltrum01@homer.louisville.edu 2Supported by NIH Grant HD29888.