In: Statistics and Probability
Two judges rate 10 contestant in a beauty contest.Test the hypothesis that the ratings are independent of one another versus the alternative that the judges tend to agree in th judgment. ( at 5% level of significance)
A B C D E F G H I J
Judge 1 1 2 3 4 5 6 7 8 9 10
Judge 2 2 3 1 4 6 5 9 10 8 7
A chi square test of independence is always a right tailed test.
The Observed and expected value tables are as below.
Each Expected Value = (Row Total * Column Total) / N, Where N = Total Observations
Observed | |||||||||||
A | B | C | D | E | F | G | H | I | J | Total | |
Judge 1 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 55 |
Judge 2 | 2 | 3 | 1 | 4 | 6 | 5 | 9 | 10 | 8 | 7 | 55 |
Total | 3 | 5 | 4 | 8 | 11 | 11 | 16 | 18 | 17 | 17 | 110 |
Expected | |||||||||||
A | B | C | D | E | F | G | H | I | J | Total | |
Judge 1 | 1.5 | 2.5 | 2 | 4 | 5.5 | 5.5 | 8 | 9 | 8.5 | 8.5 | 55 |
Judge 2 | 1.5 | 2.5 | 2 | 4 | 5.5 | 5.5 | 8 | 9 | 8.5 | 8.5 | 55 |
Total | 3 | 5 | 4 | 8 | 11 | 11 | 16 | 18 | 17 | 17 | 110 |
The Hypothesis:
H0: The ratings are independent of the Judges. (They do not agree)
Ha: The ratings are independent of the Judges. (They Tend to agree)
_______________________________________________________________
(b) The Test Statistic:
Observed | Expected | (O - E)2 | (O-E)2/E |
1 | 1.5 | 0.25 | 0.166667 |
2 | 1.5 | 0.25 | 0.166667 |
2 | 2.5 | 0.25 | 0.1 |
3 | 2.5 | 0.25 | 0.1 |
3 | 2 | 1 | 0.5 |
1 | 2 | 1 | 0.5 |
4 | 4 | 0 | 0 |
4 | 4 | 0 | 0 |
5 | 5.5 | 0.25 | 0.045455 |
6 | 5.5 | 0.25 | 0.045455 |
6 | 5.5 | 0.25 | 0.045455 |
5 | 5.5 | 0.25 | 0.045455 |
7 | 8 | 1 | 0.125 |
9 | 8 | 1 | 0.125 |
8 | 9 | 1 | 0.111111 |
10 | 9 | 1 | 0.111111 |
9 | 8.5 | 0.25 | 0.029412 |
8 | 8.5 | 0.25 | 0.029412 |
10 | 8.5 | 2.25 | 0.264706 |
7 | 8.5 | 2.25 | 0.264706 |
Total | 2.775609 |
as found above = 2.78
The degrees of freedom, df = (r – 1) * (c -1) = (2 - 1) * (10 - 1) = 1 * 9 = 9
The Critical Value: The critical value at = 0.05 , df = 9
critical = 16.92
(e) The p value: The p value at = 2.78, df = 9, is P value = 0.9724.
The Decision Rule: If test is > critical, then Reject H0.
If p value is < , Then Reject H0.
The Decision: Since test (2.78) is < critical (16.92), We Fail to reject H0.
Since p value (0.9724) is > (0.05), We Fail To Reject H0.
The Conclusion: There is insufficient evidence at the 95% significance level to conclude that the ratings are independent of the Judges.