Question

3. Many members of social media outlets like Facebook lament the fact that their friends use their Facebook page as a platform for espousing their political beliefs. Given this context, a researcher was interested to see if Facebook use varied based on an individual’s political affiliation. Specifically, he wanted to know if the average number of logins to Facebook per day varied among the various political groups. To this end, he created a survey where he asked people to indicate their political affiliation as Democrat, Republican or Other. Additionally, he asked people to indicate how many times per day they logged into/checked their Facebook account. The table below contains the Facebook use data from the 30 study participants.

Democrat	Republican	Other
6	1	4
7	3	3
10	2	1
10	5	5
4	1	3
3	1	0
1	0	3
5	3	7
6	6	4
1	2	5

1. What statistical test will you run on these data?

2. What are the assumptions of your chosen statistical test?

3. Use SPSS to test the assumptions of your chosen statistical approach. Report the results of your tests of assumptions. Have the assumptions been met?

4. Is there an overall statistically significant difference in Facebook use among the 3 political groups? How do you know?

Answer 1

a. As we see, here are three groups and the objective is to compare the mean number of logins. As we have to check the equality of more that 2 groups, we go for the one way ANOVA technique.

b. Assumptions

1. Homoscedasticity of variances among the groups

2. Dependent variables should come from normal or approximately normal population

3. No significant outliers

c. Check for the assumptions

No outliers

Check for homoscedasticity

As we see, the p-value is 0.202 which is greater than 0.05, we accept the hypothesis that there is homoscedascity.

Check for normality

We can see that the observations comes under normality.

d. By performing the test using SPSS, we can have the following outcome.

As we see the p-value is 0.041 which is less than the level of significance 0.05. Hence we reject the null hypothesis and conclude that there significance difference in Facebook use among the three political parties.