In: Math
Appendix Four:
Region
Supporter North East South West Midwest
1 23 36 45 22 40
2 66 50 48 51 75
3 43 40 29 52 65
4 70 28 30 31 30
5 60 26 38 35 67
6 49 34 87 65 78
7 54 59 20 19 67
8 64 51 26 29 76
9 54 50 39 43 44
10 60 42 38 54 61
11 70 58 50 48 70
12 49 48 29 18 90
13 38 37 21 34 78
14 40 31 29 30 69
15 49 52 40 50 28
One-way ANOVA: Age versus Region
Source | DF | SS | MS | F | P |
Region | 4 | 6612 | 1653 | 7.50 | 0.000 |
Error | 70 | 15418 | 220 | ||
Total | 74 | 22030 |
S = 14.84 R-Sq = 30.01% R-Sq(adj) = 26.01%
Comment: The p-value for the F-test at the above ANOVA table is 0.000 and less than 0.05 level of significance. Hence, we can conclude that at least one mean age of a region is significantly different from the remaining region at 0.05 level of significance. So, this accepts the differences in the mean ages of the supporters of the candidate based on their residencies
From the posthoc test, the differences in the mean ages are happening between the regions: East & Midwest, Midwest & South, and Midwest & West.
The p-value for Bartlett's Test statistic is 0.314 and less than 0.05 level of significance. Hence, we can accept the homogeneity of variances of ages between the regions at the 0.05 level of significance.