Published online by Cambridge University Press: 20 July 2010
Ten narrative studies involving family histories of 262 children of gay fathers and lesbian mothers were evaluated statistically in response to Morrison's (2007) concerns about Cameron's (2006) research that had involved three narrative studies. Despite numerous attempts to bias the results in favour of the null hypothesis and allowing for up to 20 (of 63, 32%) coding errors, Cameron's (2006) hypothesis that gay and lesbian parents would be more likely to have gay, lesbian, bisexual or unsure (of sexual orientation) sons and daughters was confirmed. Percentages of children of gay and lesbian parents who adopted non-heterosexual identities ranged between 16% and 57%, with odds ratios of 1.7 to 12.1, depending on the mix of child and parent genders. Daughters of lesbian mothers were most likely (33% to 57%; odds ratios from 4.5 to 12.1) to report non-heterosexual identities. Data from ethnographic sources and from previous studies on gay and lesbian parenting were re-examined and found to support the hypothesis that social and parental influences may influence the expression of non-heterosexual identities and/or behaviour. Thus, evidence is presented from three different sources, contrary to most previous scientific opinion, even most previous scientific consensus, that suggests intergenerational transfer of sexual orientation can occur at statistically significant and substantial rates, especially for female parents or female children. In some analyses for sons, intergenerational transfer was not significant. Further research is needed with respect to pathways by which intergenerational transfer of sexual orientation may occur. The results confirm an evolving tendency among scholars to cite the possibility of some degree of intergenerational crossover of sexual orientation.