none" to "practically all." (4)Pers

none" to "practically all." (4)Personal contact was assessed by asking respondents how many of their close female friends during the past two years were lesbian and whether their past interactions (if any)with lesbians were generally positive or generally negative. The items were repeated in reference to gay men. (5)Religiosity was assessed through three measures: frequency of attendance at religious services, the orthodoxy sub scale of the Religious Ideology Scale, or RIS (adapted from Putney & Middleton, 1961), and conservatism of respondent's denomination (dichotomized as fundamentalist vs. liberal/no religion from a continuum described by Paige, 1977).The reliability (alpha) coefficients for the scales were all greater than.70, with the exception of the ambiguity tolerance scale (alpha = .59).Because of this lower-than-optimal degree of internal consistency, the role of ambiguity tolerance in attitudes toward homosexuality must be interpreted with caution in the present study.Sample and Procedure The series of studies in the present research employed convenience samples of college undergraduates. The limits of generalizability of findings with such samples are fully recognized by the author. Never theless, given the nascent stage of empirical research on attitudes toward lesbians and gay men, such data can be useful in formulating hypotheses for later testing with more representative samples.Consequently, the data reported here should be viewed as preliminary and suggestive rather than definitive.A sample of 368 undergraduate volunteers at a major California university (249 females, 119 males) completed the ATLG scale during a lecture period in introductory psychology courses.
2 One hundred thỉrty nine of these students subsequently volunteered to participate in additional research for extra credit points. Three weeksafter initial-ly completing the ATLG, they completed a battery of paper-and-pencilmeasures which included the independent measures and alternateversions of the ATL and ATG sub scales.The parallel form of the ATLG was included in order to address thequestion of whether individuals' attitudes toward lesbians are morenegative or positive overall than are attitudes toward gay men.Alternate versions of the scales were necessary because raw mean scores from the two different scales are not directly comparable. The parallel versions were created by rewriting the ATG items so that they 2 As a check for item-order effects, one-half of the sample completed the ATLG with items in reversed order. No significant differences in mean scores were noted betweenthe two versions