In: Statistics and Probability
Examine the table below which estimates the number of voters by race in the 2016 election in a particular county. Do the results from the election suggest President Trump earned the expected number of votes from each racial/ethnic group based on their distribution in the county’s population? Performs an appropriate hypothesis test to examine this issue, being sure to list all steps and writing out your conclusion in words. (HINT: Did Trump’s vote proportion match those of the proportion of residents of each racial group)
% of Residents |
Trump (2016) |
|
African-American/Black |
16.6 |
241 |
Asian |
13.9 |
316 |
Hispanic/Latino |
17.0 |
169 |
Other |
3.2 |
68 |
White |
49.3 |
1464 |
Ans:
Chi square test for goodness of fit:
Percentage | Observed(fo) | Expected(fe) | (fo-fe)^2/fe | |
African-American/Black | 16.60% | 241 | 374.828 | 47.78 |
Asian | 13.90% | 316 | 313.862 | 0.01 |
Hispanic/Latino | 17% | 169 | 383.86 | 120.26 |
Other | 3.20% | 68 | 72.256 | 0.25 |
White | 49.30% | 1464 | 1113.194 | 110.55 |
Total | 100.00% | 2258 | 2258.00 | 278.86 |
Test statistic:
Chi square=278.86
df=5-1=4
p-value=CHIDIST(278.86,4)=0.0000
Reject the null hypothesis.
The sample data does not suggest that President Trump earned the expected number of votes from each racial/ethnic group based on their distribution in the county’s population.