Browsing resource, all submissions are temporary.
This problem is essentially the same as the F-test problem in Week 11, but you will do the non-parametric analogue of the tests.
You will analyze the political survey data from Stat 100 in Spring 2016. The data can be downloaded here and then loaded to R using the command
stat100 <- read.csv("Stat100_Spring2016_Laska_Survey03x.csv")
The variable in each column is described on this webpage.
In one survey question, students are asked to rate how important they thought "gun control" was on a scale of 0-10 (with 0 meaning not at all important and 10 meaning extremely important). We want to see if there are any differences between students from a small town, a medium-sized city (like Champaign Urbana), a big city suburb (like Chicago suburb) or from a big city (like Chicago, but not the suburbs). (It is useful to make box plots of rank(gunControl) for these groups before attempting the questions below).
a. Perform a Kruskal-Wallis test to see if there are any differences on student's attitudes towards gun control among students from different home towns. Enter the p-value to 3 significant figures.
P-value =
What do you conclude? That none of the group's attitude towards gun control is significantly different from each other in the population. That all of the groups' attitudes towards gun control are significantly different from each other in the population. That at least one of the group's attitude towards gun control is significantly different from the others in the population.
b. Perform pairwise Wilcoxon tests with Bonferroni correction to adjust the p-values. Determine from the result which pairs of groups show significant differences. Select all that apply. "Big City" - "Big City Subrub" "Big City" - "Medium City" "Big City" - "Small Town" "Big City Subrub" - "Medium City" "Big City Subrub" - "Small Town" "Medium City" - "Small Town"
