Date of Award

Summer 2019

Document Type

Open Access Dissertation



First Advisor

Bret Kloos


Homeless individuals are often the targets of negative stereotypes and significant stigmatization, which can contribute to restrictive and punitive approaches to ending homelessness. Many researchers and policymakers have sought to understand and change attitudes toward the homeless to allow for consideration of a broader range of responses to addressing homelessness. Despite attention paid to understanding attitudes, a lack of reliable and valid measurement creates methodological barriers to assessing people’s attitudes and comparing those attitudes across studies and populations.

Attitudes toward homeless persons have been demonstrated to be quite complex, which has likely impeded the development of valid and reliable measurement tools. There is a need to identify those elements of psychological theory that can best represent people’s complex attitudes toward homeless individuals. This dissertation proposed a new theoretical framework for understanding attitudes toward homeless people by integrating four theories: stereotype content model, dehumanization, attribution theory, and integrated threat theory. This model was used to inform item development for an assessment tool that reliably measures attitudes toward homeless persons.

This dissertation aimed to develop a valid and reliable measure of cognitive attitudes toward homeless people that clarifies the complexities of attitudes toward the homeless. It was conducted in three studies. Study 1 included a factor analysis of a large dataset (n = 899) from community random digit dial surveys to examine the psychometric properties of the most commonly used measure of attitudes toward homelessness, the Attitudes toward Homelessness Inventory (Kingree & Daves, 1997). Study 2 developed and tested a large item pool based on the constructs proposed in the theoretical framework. The item pool was pilot tested online (n = 2105). The best performing items were selected to create the one-factor Survey of Attitudes toward Homeless People (SAHP). In Study 3, confirmatory factor analysis was used to verify the one-factor structure and item fit in a new online sample (n = 824). In addition, construct validity and test-retest reliability was examined to establish the SAHP’s nomological network and to examine stability. The final 9-item measure demonstrated excellent internal consistency, strong test-retest reliability at 9 months, and strong construct validity (i.e., strong associations with intergroup disgust sensitivity, intergroup anxiety, blame, anger, pity, help, danger, fear, avoidance, segregation, coercion, past contact with homeless individuals, and germ aversion). The new measure offers a more reliable and more theoretically-based assessment of attitudes toward homeless individuals, which may afford greater personalization of interventions targeting public attitude change.