Campus Safety Report

Dear writer this paper was assigned in CJ Data Analysis class. It requires knowledge in SPSS by IBM. Please read closely the instructions. I will include the Labs which the professor refers to for a reference if you might need it THEY DO NOT NEED TO BE COMPLETED! Cross-tabulations need to be included into the report. If you may have any questions please contact me as soon as possible as we would avoid delay in the paper. Assignment 2 Since incidents such as the shooting at Virginia Tech and other college campuses, UML has been increasingly concerned with the safety and well-being of its students. The Chancellor has given the Department of Criminal Justice & Criminology funds to conduct a survey of students to examine rates of victimizations amongst the student body, as well as query students on their perceptions of various issues related to campus safety. As such, the Department launches a web-based survey entitled the Campus Safety Survey during the month of April of 2008 and surveys all students who were actively enrolled in classes for both the Fall 2007 and Spring 2008 semesters (n=6368). Students were invited to partake in the survey via a message to their student e-mail which included an embedded web link to the survey. After all was said and done, 26% of UML students responded to the survey (n=1670). The survey asked a number of questions about perceptions of campus safety and also whether the respondents were victimized by one of eight crimes in the previous (Fall 2007) semester: assault, sexual assault, stalking, hate crime, verbal harassment, property vandalism/theft, break-ins, and identity theft. The survey also asked a number of questions regarding the demographic characteristics of respondents, as well as illegal drug use. (Check the codebook on Blackboard for the actual questions) With the survey finished, it is up to you to analyze the data and write up a report to Chancellor, which she will distribute to other key members of the university. The Chancellor is concerned with a few key issues: (1) how many students have been victimized by crime during the Fall 2007 semester? (2) are certain kinds of students more or less likely to be victimized?; (3) what are the outcomes for people who experience victimizations? Your assignment is to write the report. It should be no more than 6 pages in length. To successfully complete this report, you must do the following: • Describe the key questions you intend to address in the report, what analyses you are undertaking, and why • Describe the survey and its methodology (as above) • Describe your measures: how were the key variables operationalized? • Prepare frequency and crosstabluation tables, with corresponding Chi-Square tests, for each of your variables • Present and discuss the results of your analyses • Provide a conclusion based on your analyses The report does not need to be in the same format as your research brief. You can write the report as you would any other paper: double-spaced with traditional font size and margins. Discussion of Assignment 2 The data contain eight different types of victimizations, which is too many to analyze individually for a short report to the Chancellor. Instead, we’re going to create a composite measure (see Lab 8) which will be our victimization variable. This variable will measure whether a student said they were a victim for any one or more of the victimization variables (so it will simply dichotomize victimization into yes and no). This will allow us to answer question 1, which will simply involve running a frequencies on our newly constructed victimization variable, and will be the dependent variable (DV) for question 2. In terms of explaining which students are more likely to be victimized, we’re going to look at several independent variables (IV) taken from the survey regarding basic demographic information from each respondent: age, race, gender, live on or off campus, class standing, and prior drug use. Some of these variables are simple and ready to use (e.g., gender, coded as male or female), but others are not. The age variable is ratio-level, so we’ll need to recode it into categories (see the Assignment 2 To-Do List). The race variable has 7 categories, and some of these categories have very few respondents, so we’ll recode it as white and non-white (also see To-Do List). Finally the drug variable has quite a large number of response categories, and is also a “check all that apply” type question, so we’ll need to create a composite measure for this as well (see Lab 8). We’ll create a variable which will simply measure whether or not the student admitted to illegal drug use during the reference period. That will get the data in shape to run our analyses, which will be the IVs of age (recoded), race (recoded), gender, on/off campus, class standing, and drug use (composite measure) and their impact on victimization. We can do this by examining their cross-tabulations and Chi-square tests (see Lab 7). In terms of explaining the outcomes of victimizations, the survey contains three items we might reasonably expect to be different for those students who have been victimized and those who have not. These are: feelings of overall safety on campus, fear of a future victimization on campus, and whether or not fear of crime has kept a student from attending campus events and activities. We might expect those who have been victimized to feel less safe on campus, to be more fearful of a future victimization, and that fear of crime has had a deeper impact on their campus activities. Again, we can test these hypotheses using cross-tabulations and Chi-square tests. Since each of these questions has only a few response categories, and we already have a victimization variable, no data prep is needed. We can simply run the analyses. Notice here though were looking at the effects of victimization on 3 different variables of interest. So, for these last 3 cross-tabulations, victimization becomes the IV. This means that it will be the column and not the row variable! Once all of these 9 cross-tabulations and their corresponding Chi-square tests are done, we need to see which are statistically significant. We don’t want to bother reporting the non-significant results, because what are seen in those tables is due to sampling error and support the null hypothesis (which is that the IV has no effect on the DV). Once we know which ones are significant, we need to format the SPSS tables to be more viewer-friendly and put them in our report (see Lab 9). See the “Writing the Report” handout for more guidance on what to put in the report.