The Development and Validation of the Sexual Prejudice Scale (SPS)

Background. Traditional measurement of antigay bias has focused on older notions of the problem (i.e., homophobia) and failed to demarcate the source of prejudice. With that aim, this study presents and evaluates a newly developed multidimensional measure of antigay bias. Items for the Sexual Prejudice Scale (SPS) reflect contemporary conceptualization of antigay bias and are based on a well researched model of prejudice. The SPS was designed for social work students; hence, the items are subtle because social work students are exposed to the values of the profession that advance equality for oppressed populations. While the acquisition of prejudice is multimodal, its chief components remain stable—disparaging cognitions and negative affective reactions. Measurement of latent phenomenon is challenging, but is improved when a clear theoretical delineation of the construct is followed by operationalization. Applying this approach, the SPS was designed to reflect the two primary forms of sexual prejudice—biases against gay men and those against lesbians.

Methods. The SPS was subjected to a three stage process to establish evidence of validity. A content validation study was undertaken to delimit the initial item pool. Data were then collected from students (n = 851) at four geographically diverse universities with accredited social work programs. The paper-and-pencil surveys were collected during regular class periods, and the instrument package included the SPS, standardized scales, and a demographic questionnaire. The dataset was split to generate two samples. An exploratory factor analysis (EFA) was used with the first sample to test the hypothesized factor structure. The second dataset was used to test for evidence of factorial validity (a confirmatory factor analysis; CFA), construct validity, known groups, and predictive validity along with a reliability analysis.

Results. Data met the assumptions of the analyses, though 15 of the 82 items were removed due to skew and kurtosis. Correlation analysis revealed that all items were correlated with at least one other item ( > .30), but three inter-item correlations surpassed .80. Three items were removed. Results of the EFA for the gendered scales indicated a three factor solution (affective-valuation, stereotyping, and social equality beliefs) with correlated factors, which was conceptually consistent with the model for sexual prejudice. Bartlett's Test of Sphericity and the Kaiser-Meyer-Olkin measures were excellent. A three factor model for the gendered scales was supported by CFA as evidenced by the goodness-of-fit indices. Cronbach's alphas for each subscale ranged from adequate to good while the composite scales' stratified alphas were excellent. A priori validity hypotheses were confirmed.

Conclusion. Evidence of validity and reliability support the SPS usefulness with social work students. The scale can be used to test established pedagogical interventions, which will allow researchers to determine how the components of prejudice are addressed by these approaches. In addition to its applicability for research and intervention development, the SPS can be used to increase individual awareness about prejudice and promote a discussion about how prejudice is related to client harm. Future research seeks to test the utility of the SPS with other populations, such as practicing social workers.