a bipolar Likert scale)? How many police officers do you know personally? Using the scale (semantic differential), how would you describe police officers? Semantic differential scales are used to measure perception of a group or object by using adjectives on a bipolar scale, for example, participants would rank police officers on a scale ranging from polite to impolite. The survey resulted in a total of 67 responses, which we then analyzed.

Once we gathered the data from the survey, we coded the responses so that we would be able to see the relationships between factors more easily. For example, for a yes or no question, group 1 would be the participants that responded "yes" and group 2 would be the participants that responded "no". Then, we had to assess the relationship between certain factors, like sex and their perception of the police, for example, so we conducted a t-test. T-tests were also conducted to determine the relationship between exposure to the criminal justice system and perception of police and to assess the relationship between political affiliations with perception of police. ANOVA was used to assess the relationship between location and perception of police. A correlation was performed to show whether an individual's relationship with police affected their perception of police.

When we began the research, we thought about the effect location would have on perception of the police. We hypothesized that individuals living in urban areas would have a higher score (more negative feelings towards police) than suburban residents, and that suburban residents would then have a higher score than those living in rural areas. This hypothesis arose from the observation that urban areas have more crime than rural areas so the urban residents would be more likely to have more (possibly negative) interactions with the police. This hypothesis was supported by our results. We also hypothesized that the number of relationships a participant had with a police officer would correlate with their perception. We believed this would affect perception significantly because participants would likely see police officers in a more positive way (a lower score) if they had family members, for example, who were police officers. We conducted a bivariate correlation using Pearson's Product Moment to test our hypothesis.

This hypothesis was also supported by our data; the number of relationships the participant had with the police negatively correlated with the average of scores, r(52) = -