In: Statistics and Probability
Compute chi square and make a determination about the null hypothesis.
Expected Frequencies | Men | Women | Total |
Agree | 269.2 | 316.8 | 586.0 |
Disagree | 233.8 | 275.2 | 509.0 |
Total | 503.0 | 592.0 | 1095.0 |
Observed Frequencies | Men | Women | Total |
Agree | 242.0 | 344.0 | 586.0 |
Disagree | 261.0 | 248.0 | 509.0 |
Total | 503.0 | 592.0 | 1095.0 |
(Observed-Expected)2 / Expected | Men | Women | Total |
Agree | 2.8 | 2.3 | 585.0 |
Disagree | 3.2 | 2.7 | 509.0 |
Total | 503.0 | 592.0 | 1095.0 |
Solution :
To test variables Gender(male,female) and response(agree,disagree) are independent or not.
The hypothesis are,
Null hypothesis Ho : The variables gender and response are independent.
Vs
Alternative hypothesis Ha : The variables gender and response are not independent.
Given that ,
Observed Frequency | Expected Frequency | (O-E)^2/E |
242 | 269.2 | 2.8 |
344 | 316.8 | 2.3 |
261 | 233.8 | 3.2 |
248 | 275.2 | 2.7 |
So by taking the sum of values in the last column, we get the value of chi-square test statistics.
Now using chi-square table with degrees of freedom (r-1)*(c-1) = (2-1)*(2-1) = 1 (where r=rows and c=columns) and level of significance = 0.05 , we get critical value 3.84.
Decisiosn rule : Reject Ho ,if test statistics > critocal value otherwise fail to reject Ho.
So here as Test statistics(11) > critical value ,we reject Ho.
Conclusion : The variables gender and response are not independent.