The aim of this study is to explain the effect of arousal on consumer’s attitude towards a direct service in terms of evaluation extremity. An important element within consumer behavior is the attitude that consumers have towards products and services. After the conceptual framework was developed through an extensive literature review on arousal, attitude and services, a hypothesis was formulated stating that: Arousal has an effect on consumer's attitude towards a direct service in terms of their evaluation extremity.  An experiment was carried out within various lectures of a university in Sweden, whereas the sampling used was a convenience sample of university students. The total amount of respondents who participated in the experiment were 154, whereof 102 were valid respondents. The experimenttriggerused to influence their level of arousal was caffeinated coffee, while the control trigger was decaffeinated coffee. The Fishbein model was used to measure attitude translated into a questionnaire which was administered at the experiment, while the three adjective scale was utilized to measure the level of arousal for control. The collected data was imported into the statistics program SPSS for analysis and discussion. Quality criteria and ethics standards were adhered to  during the entire research process. At the conclusion of the research, the hypothesis was accepted. However, it should be taken into consideration that the caffeine did not have the desired effect on the level of arousal, and the study mainly used the control questions for arousal to find a significance level. The study was deemed successful though, in that after calculating positive pretests for validity, the results showed that the aroused participants had a significantly more extreme attitude than the non-aroused participants. In that the experiment and results are original, it allows for further related research. Finally, to insure the validity of the results of this study, it is pertinent that the experiment’s results are replicated by future studies.

Four studies investigated the effects of attitude extremity and political ideology on the degree and direction of changes in issue attitudes following the presentation of mixed evidence. Based upon previous work, it was predicted that those holding relatively more extreme attitudes would resist changing those views when presented with a mixture of supporting and opposing statements and would potentially adopt more extreme evaluative positions – a phenomenon known as attitude polarization (Lord, Ross, & Lepper, 1979). Evaluative entrenchment or intensification was also expected among more politically conservative participants, based upon prior work describing cognitive rigidity and resistance to change as more characteristic of the political right than left (e.g., Jost, Glaser, Kruglanski, & Sulloway, 2003). An interaction of attitude extremity and political ideology was also hypothesized, such that liberal individuals with moderate attitudes were expected to demonstrate the least propensity to polarize. Participants’ attitudes regarding abortion rights (Study 1), gun control (Study 2), tax increases (Study 3), and environmental preservation (Study 4) were assessed before and after reading statements that both opposed and supported the issue. Political ideology was also assessed, along with several individual difference factors. Across all four studies, attitude extremity significantly predicted evaluative change, although the pattern of that effect varied. Political ideology did not emerge consistently as a predictor of attitude change; however, significant interactive effects of extremity and ideology were found. In addition, several individual difference factors (i.e., gender, need for cognition, issue importance) were found to moderate the effects of the primary predictors on attitude change, and some divergent result patterns were found when comparing data from a college and non-college sample in Study 4. Taken together, these studies provide evidence that attitude extremity and political ideology influence the degree and direction of evaluative change following the presentation of mixed evidence. In addition, they identify other factors at work in effecting change versus resistance, thereby highlighting the multi-faceted and complex nature of persuasion in a political context.