Two Population Proportion Inference
Data Summary
Counts: Raise Cigarette Tax by Smoking
| - |
Yes |
No |
Total |
| Non-Smoker |
351 |
254 |
605 |
| Smoker |
95 |
100 |
195 |
Two Population Proportion Test of Hypothesis
Method: Large Sample z Test (Pooled Standard Error)
Success = Yes
Population 1 = Non-Smoker, Population 2 = Smoker
Sample Size: Non-Smoker = 605, Smoker = 195
Number of Successes: Non-Smoker = 351, Smoker = 95
Proportion of Success: Non-Smoker = 0.5802, Smoker = 0.4872
Significance level = 5%
Alternative Hypothesis Ha: Proportion of 'Non-Smoker -
Smoker' is greater than 0
| Proportion Non-Smoker |
Proportion Smoker |
Difference |
Std Error |
Obs z Stat |
5% z-Upper Critical |
P-Value |
BFB |
| 0.580165 |
0.487179 |
0.0929858 |
0.0409005 |
2.27346 |
1.64485 |
0.0114992 |
7.16424 |
- Test is significant at 5% level.
-
Bayes Factor Bound (BFB): The data imply the odds in favor of
the alternative hypothesis is at most 7.16 to 1, relative to the
null hypothesis.
P-value Graph: Large Sample z, Pooled SE
Null density (in units of data): Normal; mean = 0 , sd = 0.040901
Alternative Hypothesis Ha: Proportion of 'Non-Smoker - Smoker' is
greater than 0
