In: Statistics and Probability
a professor is teaching three sections of the same course and wants to determmine if the proprtion of students who passed is statistically different across the classes. the report below shows how many students in each section passed and how many were enrolled
Section 1 - 82 passed 107 enrolled
Section 2 - 74 passed 108 enrolled
Section 3 - 89 passed 115 enrolled
conduct the chi-square test and report the final test statistic
null hypothesis: Ho: Proportion of students who passed is same across the classes
Alternate hypothesis:Ha:Proportion of students who passed is statistically different across the classes
degree of freedom(df) =(rows-1)*(columns-1)= | 2 |
for 2 df and 0.05 level of significance critical region χ2= | 5.991 |
applying chi square test:
Expected | Ei=row total*column total/grand total | section 1 | section 2 | section 3 | Total |
Passed | 79.44 | 80.18 | 85.38 | 245 | |
failed | 27.56 | 27.82 | 29.62 | 85 | |
total | 107 | 108 | 115 | 330 | |
chi square χ2 | =(Oi-Ei)2/Ei | section 1 | section 2 | section 3 | Total |
Passed | 0.0825 | 0.4766 | 0.1536 | 0.713 | |
failed | 0.2379 | 1.3737 | 0.4427 | 2.054 | |
total | 0.320 | 1.850 | 0.596 | 2.767 |
test statistic X2 =2.767
as test statistic is not in rejection region we can not reject null hypothesis
we do not have evidence at 0.05 level to conclude that Proportion of students who passed is statistically different across the classes