Negative attitudes held by neurotypical co-workers are often a major barrier to employment for IAS (Griffith, Totsika, Nash, & Hastings, 2011). Roskos-Ewoldsen and Fazio (1997) demonstrated that accessible beliefs are more predictive of people’s attitudes. The purpose of this dissertation was two-fold. First, it sought to disentangle the extant research on the relationship between attitudes and beliefs. Second, it aimed to test the effectiveness of a theoretically-based narrative intervention designed to bolster pre-existing positive and change pre-existing negative attitudes toward working alongside individuals on the autism spectrum by increasing the accessibility of certain beliefs. Study 1 was of a 2 (1 v. 4 belief repetitions) x 2 (transportation high v. low) x 3 (autism label: beginning, end, none) between-subjects experimental design. Participants (N = 555) were randomly assigned to watch a narrated video. After, participants completed a belief listing task. To determine which attitude-belief relationship model accounted for the most of variance, a hierarchical analysis was conducted. Results indicated that the modified summation model (Cronen & Conville, 1973; 1975a; 1975b) did the best job of predicting attitudes.To test the effectiveness of the narrative intervention, several hypotheses were proposed. Unfortunately, many of the hypotheses testing the direct effects of the manipulations were not supported. Two models were developed to test the relationships between two key concepts from the entertainment overcoming resistance model (EORM; Moyer-Guse, 2008). Results provided evidence supporting significant inverse relationships between both transportation and identification and reactance. Additionally, findings suggested a positive relationship between belief accessibility and attitudes. However, the lack of support for a successful manipulation impedes the opportunity to make causal inferences. Study 2 aimed to replicate support for the more parsimonious modified summation model (Cronen & Conville, 1973; 1975a; 1975b) as the best model for predicting attitudes from beliefs found in Study 1. Also, Study 2 sought to determine if motivation influenced belief accessibility. Study 2 was of a 2 (1 v. 4 belief repetitions) x 2 (motivation to process the message: high vs. low) x 2 (autism mentioned or autism not mentioned) between-subjects experimental design. As with Study 1, participants (N = 401) viewed one version of a narrated slideshow.Again, results from Study 2 garnered support for Cronen and Conville's (1973; 1975a; 1975b) model. For the relationship between motivation and attention, findings revealed that when self-involvement was increased, individuals paid more attention. However, there was no evidence of an indirect effect of motivation on attitudes via belief accessibility. Interestingly, as in Study 1, results indicated a positive relationship between belief accessibility and attitudes toward IAS and working with IAS.Theoretically, results rendered evidence for the use of narratives to reduce… Advisors/Committee Members: Ewoldsen, David (Advisor).