Homosexuals in the armed forces: a reply to Cameron, Cameron, and Proctor.

Abstract

Cameron, Cameron, and Proctor (1988) assert that current policies that exclude homosexuals from military service should be maintained, since their data purportedly demonstrate that many persons have highly negative attitudes towards homosexuals. However, their data have questionable validity and alternative implications. These authors' national sam~le clearlv is not re~resentative of the United States population. Their sample is younger, disproportionately female, more educated, disproportionately white, and more urban, with most respondents from Los Angeles, Denver, Omaha, Louisville, and Washington, DC. Furthermore, 52.5% of their potential respondents refused to be interviewed; the median age of the rejectors n7as 55 yr., whereas that of the respondents was only 34 yr., indicating that respondents and rejectors represented very different populations. Finally, data were obtained from only 842 (20.25%) of the 4,158 potential respondents in their Dallas sample. It seems unlikely that these authors' data accurately portray societal attimdes towards homosexuals. These authors' data, if they do have merit, might imply that efforts should be expended to increase societal acceptance of homosexuals rather than that sanctions against them should be maintained. Homosexuals' alleged difficulties in society and in the milimy may stem primarily from the stresses that result from negative attimdes towards them, and disciplinary personnel might be especially alert to rhe infractions of homosexuals. Such stresses and biases might account for these authors' observation that about half of the troubles into which homosexuals get in the military were not connected with their homosexuality per se. Finally, these authors' observation that majority of males diagnosed with borderline personality disorder are homosexual does not imply that a significant proportion of homosexuals display borderline personalicy disorder. In conclusion, these authors' data have serious limitations, and inferences from