Abstract
Severe deficits in face recognition, or prosopagnosia, cluster in some families. However, the degree of family resemblance (familiality) for face memory ability in the general population is unknown. Moreover, familiality can result from either genetic or environmental factors shared between family members. We conducted a classical twin study in an unselected population to quantify family resemblance in face memory ability and to parse this resemblance into genetic and environmental components. We measured face memory with the Cambridge Face Memory Test (CFMT), a highly reliable measure (Spearman-Brown corrected split-half reliability=.82) that correlates little with other memory tests or tests of general intelligence. We observed a high correlation in performance among monozygotic (MZ) twins, implying high familiality of face memory (intraclass r(69)=.75, 95% CI=.63–.84); since MZ twins share both genes and environment, the MZ twin correlation indicates total family resemblance. This MZ correlation approaches the ceiling set by the CFMT's measurement reliability, suggesting that most of the reliable variation in CFMT performance is familial. Since both MZ and dizygotic (DZ) twins share family environment, but MZ twins share twice as many genes as DZ twins (100% vs. 50%), any greater correlation in MZ than DZ twins supports an effect of genes. The DZ correlation we observed (intraclass r(21)=.52, 95% CI=.14–.77) was significantly lower than the MZ correlation (p=.05, one-tailed), evidence that genetic factors cause some differences in face memory ability. In sum, our results provide evidence that face memory ability is highly familial and at least partially genetic.