In: Math
Design A Design B Design C
16 33 23
18 31 27
19 37 21
17 29 28
13 34 25
Use the Kruskal-Wallis H test and the Chi-Square table at the 0.05 level to compare the three designs.
Design A | Rank A | Design B | Rank B | Design C | Rank C |
16 | 2 | 33 | 13 | 23 | 7 |
18 | 4 | 31 | 12 | 27 | 9 |
19 | 5 | 37 | 15 | 21 | 6 |
17 | 3 | 29 | 11 | 28 | 10 |
13 | 1 | 34 | 14 | 25 | 8 |
Total | 15 | Total | 65 | Total | 40 |
Null Hypothesis(H0):
All three samples come from identical populations.
Alternative Hypothesis(H1):
At least one sample comes from the different population than the others.
Test statistic:
Total sample size, n =15
n1 =n2 =n3 =5
Formula: Test statistic, H =[12/(n(n+1))*] - 3(n+1)
(where, i =1, 2, 3)
H =[12/(15*16)*((152+652+402)/5)] - 3(16) =(0.05*1210) - 48 =12.5
Thus, the test statistic, H =12.5
Critical value:
Kruskal-Wallis H statistic follows Chi-square distribution.
Degrees of freedom, df =Number of samples - 1 =3 - 1 =2
Significance level given =0.05
At 0.05 significance level and at df =2, the Chi-square critical value is: =5.991
Conclusion:
H: 12.5 > : 5.991
Since the test statistic is greater than the critical value, we reject the null hypothesis(H0) at 0.05 significance level.
Thus, there is a sufficient evidence to claim that at least one sample comes from the different population than the others.