In: Statistics and Probability
Block | |||||
Treatment | 1 | 2 | 3 | 4 | Treatment Mean |
Tr1 | 2 | 1 | 2 | 3 | 2 |
Tr2 | 4 | 4 | 1 | 1 | 2.5 |
Tr3 | 3 | 4 | 3 | 2 | 3 |
Block Mean | 2 | 3 | 3 | 2 | overall mean = 2.5 |
Consider the randomized block design with 4 blocks and 3 treatments given above. What is the value of the F statistic for blocks?
Block 1 | Block 2 | Block 3 | Block 4 | Sum | ||||
Tr1 | 2 | 1 | 2 | 3 | P1 = | 8 | N = | 12 |
Tr2 | 4 | 4 | 1 | 1 | P2 = | 10 | n = | 3 |
Tr3 | 3 | 4 | 3 | 2 | P3 = | 12 | k = | 4 |
G = ΣX = | 30 | |||||||
ΣX² = | 90 | |||||||
Sum | T1 = | T2 = | T3 = | T4 = | ||||
9 | 9 | 6 | 6 | |||||
SS1 = | SS2 = | SS3 = | SS4 = | |||||
2 | 6 | 2 | 2 |
SSTotal = ΣX² - (ΣX)²/N = 90 - (30)²/12 = 15
SSBetween = Σ((ΣTᵢ)²)*(1/n) - (ΣX)²/N = (9² + 9² + 6² + 6²)*(1/3) - (30)²/12 = 3
SSSubject = Σ((ΣPⱼ)²)*(1/k) - (ΣX)²/N = (8² + 10²+ 12²)*(1/4) - (30)²/12 = 2
SSError = SSTotal - SSBetween - SSSubject = 15 - 3 - 2 = 10
dfBetween = k-1 = 3
dfSubject = n-1 = 2
dfError = (k-1)*(n-1) = 6
dfTotal = N-1 = 11
MSBetween = 3/3 = 1
MSSubject = 2/2 = 1
MSError = 10/6 = 1.66667
F for Block = MSBetween/ MSError = 1/1.6667 = 0.6
Source | SS | df | MS | F |
Block | 3 | 3 | 1 | 0.6 |
Treatment | 2 | 2 | 1 | 0.6 |
Error | 10 | 6 | 1.6667 | |
Total | 15 | 11 |