In: Statistics and Probability
A questionnaire collects information from UTS students on gender and whether or not the student smokes. The resultant two-way table is shown below.
Women Men Total
Don’t smoke 153 166 319
Smoke 16 27 43
Total 169 193 362
a) We intend to test whether there is a difference in the proportion of men and women who smoke. Define the parameters, and state the Null and Alternative hypotheses
b) What proportion of men are smokers? What proportion of women are smokers? What is the difference between these proportions?
c) Using information associated with the plots on the next page o Which plot would you use to perform the test in part a)? o Let = 0.05, State your conclusion about whether there is a difference in the proportion of men and women who smoke - with a numerical reference from the appropriate plot
Women | Men | Total | |
Do not smoke | 153 | 166 | 319 |
Smoke | 16 | 27 | 43 |
Total | 169 | 193 | 362 |
This is the information that has been provided to us. This is the information collected from UTS students on gender and whether or not the student smokes.
a.
We are required to test if there is a difference in the proportion of men and women who smoke.
Since we have information from a sample, we need to draw inferences about the population from the sample. Thus, we need to if there is a significant difference in the proportion of men and women who smoke. The conclusion cannot be drawn looking at the sample information.
There is variability in the sample which we must account for.
Let us define the parameters:
Let pm be the proportion of men who smoke; pm= 27/166= 0.1626506
Let pw be the proportion of women who smoke; pw= 16/153= 0.1045752
Let p be the proportion of people who smoke (men and women); p= 43/319= 0.1347962
We will test the hypothesis that the proportions are equal.
The null hypothesis, H0: pm pw
The null hypothesis is that there is no significant difference between the proportion of men and women who smoke.
The alternate hypothesis, Ha: pmpw
The alternate hypothesis is that there is in fact a significant difference between the proportion of men and the proportion of women who smoke.
b.
The proportion of men who are smokers (in the sample), pm= 27/166= 0.1626506
The proportion of women who are smokers (in the sample), pw= 16/153= 0.1045752
The difference between these proportions, pm - pw= 0.1626506- 0.1045752= 0.0580754
c.
The level of significance for the z-test is 0.05 or 5%.
We must first find the z- statistic, and compare this statistic with the z critical value (value from the z-table). This is a two tailed test. Thus,
If the z statistic is greater than or less than the upper limit or the lowere limit (respectively) of the z critical value, then we can reject the null hypothesis, at 5% level of significance. If it is not, we will fail to reject the null hypothesis.
Under the assumption that the null hypothesis is true, the formula for calculating the z- statistic is
Z= (pm-pw-0) / sqrt(p(1-p)(1/nm+ 1/nw) )
where
pm= proportion of men who smoke
pw= proportion of women who smoke
p= proportion of total smokers
nm=
number of men
nw= number of women.
Thus, Z= (0.0580754-0) / sqrt (0.1347962* (1-0.1347962)(1/169 + 1/193)
= 0.0580754 / sqrt (0.1166262* 0.01109851)
Thus, the Z critical value is 1.614217
The value from the z-table, for a two tailed test with 5% level of significance is + or - 1.96.
Since our z statistic is less than the z critical value, we fail to reject the null hypothesis, and conclude that there is no significant difference in the proportion of men and the proportion of women who smoke.