In: Statistics and Probability
Exercise 12. A researcher claims that the proportion of students that are pursuing a Bachelor of Arts degree and must work full time matches the distribution shown in the table on the left below. To test this claim, the researcher randomly surveyed 200 students, 50 from each year of study). The results are shown table on the right. At α = 0.05, is there evidence to support the researcher’s claim that the findings match the claimed distribution?
Researcher's Claim
CLASS Work full-time
Freshman 30%
Sophomore 32%
Junior 34%
Senior 38%
Researcher’s findings (n = 50 per class)
CLASS Work full-time
Freshman 14
Sophomore 18
Junior 17
Senior 21
We have to use goodness of fit chi-square test for testing whether the researcher's claims are consistent with the observed values.
There are 4 grades with 50 students in each class. To find out the expected number of full timers we will multiply 50 * (proportion of each grade).
Eg.: Freshman Ei = 50 * 0.30 = 15
Grade |
Proportion assumed |
Eepected Ei |
Observed Oi |
|Ei-Oi| | ||
Freshman | 0.3 | 15 | 14 | 1 | 1 | 0.066667 |
Sophomore | 0.32 | 16 | 18 | 2 | 4 | 0.25 |
Junior | 0.34 | 17 | 17 | 0 | 0 | 0 |
Senior | 0.38 | 19 | 21 | 2 | 4 | 0.210526 |
Total | 0.527193 |
The findings are consistent with the researcher's claim(distribution).
VS
The findings are not consistent with the researcher's claim(distribution).
Test Statistic:= 0.527193
Critical value: where 'k' is the number of categories
( can be found online from chi-square tables or using excel func 'chiinv' )
Decision Criteria: Reject null hypothesis if T.S. > C.V.
0.527193 < 9.3484
Decision: Since T.S. < C.V,we do not reject the null hypothesis at 5% level of significance.
Conclusion: The observed values matches with the researcher's claim. Hence,the distribution is a good fit.