y are made by attempting to separate genetic from environmental sources of variance. Evidence exists that intelligence runs in families: the correlation of IQ scores of genetically related people increases by the closeness of relationship, and "this correlation pattern remains even when controlling for social class, education, race, gender, and the like (Herrnstein & Murray 1994)." Lykken, Bouchard, & McGue (1993) found an average correlation of about .45 for IQ scores of biological parents and offspring and for siblings living together. Most of our understanding of the genetics of intelligence is grounded in twin and adoption studies, which have documented significant and substantial genetic influence (Plomin, 1994, Plomin & Petrill, 1997, Steen, 1996): for example, correlation between scores of monozygotic (MZ) twins reared together is higher (approximately .85) than correlations of dizygotic (DZ) twins and less closely related siblings (Plomin & Petrill, 1997). There is also a high correlation in IQ scores between MZ twins who were raised apart (Joseph, 1998). Dudley (1991) found a higher correlation between IQ scores of adopted children and their biological parents than with their adoptive parents. A major concern with both twin and adoption studies, however, is the amount of correlation in the environments of adopting and biological families (Lewontin, Rose, & Kamen, 1984). There have also been a number of methodological problems (Steen, 1996), as well as several instances of fraud (Cyril Burt, for example). Consequently, it is difficult to use these studies to bolster arguments of heritability of intelligence.Exactly how much intelligence is attributable to genetics is unknown, and estimates vary widely. Arthur Jensen (1969, as cited in Eyesenck, 1998) placed heritability of intelligence at 80 percent, Eyesenck (1998) at 70 percent, Herrnstein and Murray (1994) between 60 and 80 percent, and Plomin and Petrill (1997) at 50 to 60 p...
