1. The null hypothesis is that fear of walking in one’s neighborhood and gun ownership are unrelated. In other words, that any relationship I find between these two variables in my sample is due to sampling error.
2. My research hypothesis is that people who are afraid to walk in their neighborhood will be more likely to own a gun (one-tailed).
3. Fear is the independent variable, and gun ownership is the dependent variable. Both of my variables are nominal-level variables (both have yes/no response options).
4. I will use a .05 p-value as my rejection /significance level.
5. The appropriate test for examining the relationship between 2 nominal variables is the chi-square.
6. See findings pasted below.
7. My p-value for this analysis was .016, which is below the .05 cut-off. This means that my
findings are statistically significant, and I can reject the null hypothesis. However, the
relationship between these variables was not in the direction I expected, and I cannot
claim support for my original research hypothesis. Interestingly, people who are afraid
to walk alone at night are LESS likely to own a gun. Specifically, 24% of people who are
afraid in their neighborhood own a gun at home, while nearly 33% of those who are not
afraid own a gun (X2 = 8.23, df=2, p = .016). Perhaps having a gun at home leads people
to feel slightly more safe.