1.1 ONE-FACTOR MODEL Y = A + e with planned orthogonal contrasts Three-level factor A. Analysis by GLM of terms: B + C(B) Data: A B C Y 1 2 0 4.5924 1 2 0 -0.5488 1 2 0 6.1605 1 2 0 2.3374 1 2 0 5.1873 1 2 0 3.3579 1 2 0 6.3092 1 2 0 3.2831 2 -1 1 7.3809 2 -1 1 9.2085 2 -1 1 13.1147 2 -1 1 15.2654 2 -1 1 12.4188 2 -1 1 14.3951 2 -1 1 8.5986 2 -1 1 3.4945 3 -1 -1 21.3220 3 -1 -1 25.0426 3 -1 -1 22.6600 3 -1 -1 24.1283 3 -1 -1 16.5927 3 -1 -1 10.2129 3 -1 -1 9.8934 3 -1 -1 10.0203 COMMENT: If A[1] is a control, and A[2], A[3] are treatment levels, contrasts B and C test for a control-versus-treatment effect, and a between-treatments effect. SS[B] + SS[C(B)] = SS[A]; likewise DF[B] + SS[C(B)] = DF[A]. Model 1.1(i) A is a fixed or random factor: Source DF SS MS F P 1 A 2 745.36 372.68 17.08 <0.001 2 S(A) 21 458.20 21.82 Total 23 1203.56 Orthogonal contrasts, with fixed factors B and C: Source DF Seq SS Adj SS Seq MS F P 1 B 1 549.39 549.39 549.39 25.18 <0.001 = A[1] versus average{A[2],A[3]} 2 C(B) 1 195.97 195.97 195.97 8.98 0.007 = A[2] versus A[3] 3 S(A) 21 458.20 458.20 21.82 Total 23 1203.56 __________________________________________________________________ Five-level factor A. Analysis by GLM of terms: B + C(B) + D(C B) + E(C B) Data: A B C D E Y 1 4 0 0 0 4.5924 1 4 0 0 0 -0.5488 1 4 0 0 0 6.1605 1 4 0 0 0 2.3374 2 -1 1 1 0 5.1873 2 -1 1 1 0 3.3579 2 -1 1 1 0 6.3092 2 -1 1 1 0 3.2831 3 -1 1 -1 0 7.3809 3 -1 1 -1 0 9.2085 3 -1 1 -1 0 13.1147 3 -1 1 -1 0 15.2654 4 -1 -1 0 1 12.4188 4 -1 -1 0 1 14.3951 4 -1 -1 0 1 8.5986 4 -1 -1 0 1 3.4945 5 -1 -1 0 -1 21.3220 5 -1 -1 0 -1 25.0426 5 -1 -1 0 -1 22.6600 5 -1 -1 0 -1 24.1283 COMMENT: This example uses one of four alternative models for the set of orthogonal contrasts B-E amongst the five levels of factor A. Model 1.1(i) A is a fixed or random factor: Source DF SS MS F P 1 A 4 1017.79 254.45 25.80 <0.001 2 S(A) 15 147.95 9.86 Total 19 1165.74 Orthogonal contrasts, with fixed factors B, C, D and E: Source DF Seq SS Adj SS Seq MS F P 1 B 1 262.82 262.82 262.82 26.65 <0.001 2 C(B) 1 297.16 297.16 297.16 30.13 <0.001 3 D(B C) 1 89.99 89.99 89.99 9.12 0.009 4 E(B C) 1 367.83 367.83 367.83 37.29 <0.001 5 S(A) 15 147.95 147.95 9.86 Total 19 1165.74 __________________________________________________________________ Doncaster, C. P. & Davey, A. J. H. (2007) Analysis of Variance and Covariance: How to Choose and Construct Models for the Life Sciences. Cambridge: Cambridge University Press. http://www.southampton.ac.uk/~cpd/anovas/datasets/ http://www.soton.ac.uk/~cpd/anovas/datasets/Orthogonal contrasts.htm