In: Statistics and Probability
1. The General Social Survey (2008) asked a random sample of people whether they agree with the following statement: A husband’s job is to earn money; a wife’s job is to take care of the home. Based on this data (displayed below), conduct a chi-square hypothesis test to assess whether there are statistically significant gender differences in feelings about gender roles ( = 0.05).
Male |
Female |
Total |
|
Agree |
181 |
195 |
376 |
Neither |
142 |
145 |
287 |
Disagree |
290 |
413 |
703 |
Total |
613 |
753 |
1,366 |
a. Write out the hypotheses
b. Calculate degrees of freedom .
c. Construct a table of expected frequencies (fe).
d. Compute the chi-square statistic. Create a table like the ones we used in class to help you.
e. Determine the p-value using the appropriate Appendix.
a.
Null Hypothesis(H0):
Gender and feelings about gender roles are independent, i.e., there are statistically no significant gender differences in feelings about gender roles.
Alternative Hypothesis(H1):
Gender and feelings about gender roles are not independent, i.e., there are statistically significant gender differences in feelings about gender roles.
b.
No. of columns, c =2 (Male, Female)
No. of rows, r =3 (Agree, Neither, Disagree)
Degrees of freedom, df =(c-1)(r-1) =(2-1)(3-1) =2
c.
Table of expected frequencies (fe):
Formula: fe =(corresponding row total*corresponding column total)/Overall total
Male | Female | Total | |
Agree | (376*613)/1366 =169 | (376*753)/1366 =207 | 376 |
Neither | 129 | 158 | 287 |
Disagree | 315 | 388 | 703 |
Total | 613 | 753 | 1366 |
d.
Computation of the chi-square statistic:
Observed frequencies: fo | Expected frequencies: fe | (fo - fe)2/fe |
181 | 169 | 0.8521 |
142 | 129 | 1.3101 |
290 | 315 | 1.9841 |
195 | 207 | 0.6957 |
145 | 158 | 1.0696 |
413 | 388 | 1.6108 |
=1366 | =1366 | =7.5224 |
Chi-square statistic, =7.5224
e.
For the test statistic, =7.5224 at df =2, the p-value =0.0233
Conclusion:
Since p-value: 0.0233 < 0.05 significance level, we reject the null hypothesis(H0) at 5% significance level.
Thus, we have a sufficient evidence to claim that there are statistically significant gender differences in feelings about gender roles.