data beads; input size time; cards; 1 100 1 93 1 140 2 105 2 45 2 85 3 35 3 40 3 37 ; proc plot; plot time * size / vpos=19 hpos=50; proc glm; class size; model time=size; means size; estimate 'svl' size 1 0 -1; estimate 'svm+l' size 2 -1 -1/divisor=2; contrast 'svl1' size -1 0 1; means size /bon cldiff alpha=.15; means size/tukey cldiff alpha=.10; means size/dunnett('3') cldiff alpha=.10; means size/dunnett1('3') cldiff alpha=.10; means size/scheffe cldiff alpha=.10; run; quit; Plot of TIME*SIZE. Legend: A = 1 obs, B = 2 obs, etc. 150 + |A | | | | A 100 +A |A | A TIME | | | 50 + | A A | B | | | 0 + -+-----------------------+-----------------------+- 1 2 3 SIZE General Linear Models Procedure Level of -------------TIME------------ SIZE N Mean SD 1 3 111.000000 25.3574447 2 3 78.333333 30.5505046 3 3 37.333333 2.5166115 Bonferroni (Dunn) T tests for variable: TIME NOTE: This test controls the type I experimentwise error rate but generally has a higher type II error rate than Tukey's for all pairwise comparisons. Alpha= 0.15 Confidence= 0.85 df= 6 MSE= 527.5556 Critical Value of T= 2.44691 Minimum Significant Difference= 45.889 Comparisons significant at the 0.15 level are indicated by '***'. Simultaneous Simultaneous Lower Difference Upper SIZE Confidence Between Confidence Comparison Limit Means Limit 1 - 2 -13.22 32.67 78.56 1 - 3 27.78 73.67 119.56 *** 2 - 1 -78.56 -32.67 13.22 2 - 3 -4.89 41.00 86.89 3 - 1 -119.56 -73.67 -27.78 *** 3 - 2 -86.89 -41.00 4.89 Tukey's Studentized Range (HSD) Test for variable: TIME NOTE: This test controls the type I experimentwise error rate. Alpha= 0.1 Confidence= 0.9 df= 6 MSE= 527.5556 Critical Value of Studentized Range= 3.558 Minimum Significant Difference= 47.187 Comparisons significant at the 0.1 level are indicated by '***'. Simultaneous Simultaneous Lower Difference Upper SIZE Confidence Between Confidence Comparison Limit Means Limit 1 - 2 -14.52 32.67 79.85 1 - 3 26.48 73.67 120.85 *** 2 - 1 -79.85 -32.67 14.52 2 - 3 -6.19 41.00 88.19 3 - 1 -120.85 -73.67 -26.48 *** 3 - 2 -88.19 -41.00 6.19 Dunnett's T tests for variable: TIME NOTE: This tests controls the type I experimentwise error for comparisons of all treatments against a control. Alpha= 0.1 Confidence= 0.9 df= 6 MSE= 527.5556 Critical Value of Dunnett's T= 2.332 Minimum Significant Difference= 43.74 Comparisons significant at the 0.1 level are indicated by '***'. Simultaneous Simultaneous Lower Difference Upper SIZE Confidence Between Confidence Comparison Limit Means Limit 1 - 3 29.93 73.67 117.41 *** 2 - 3 -2.74 41.00 84.74 Dunnett's One-tailed T tests for variable: TIME NOTE: This tests controls the type I experimentwise error for comparisons of all treatments against a control. Alpha= 0.1 Confidence= 0.9 df= 6 MSE= 527.5556 Critical Value of Dunnett's T= 1.817 Minimum Significant Difference= 34.073 Comparisons significant at the 0.1 level are indicated by '***'. Simultaneous Simultaneous Lower Difference Upper SIZE Confidence Between Confidence Comparison Limit Means Limit 1 - 3 39.59 73.67 107.74 2 - 3 6.93 41.00 75.07 Scheffe's test for variable: TIME NOTE: This test controls the type I experimentwise error rate but generally has a higher type II error rate than Tukey's for all pairwise comparisons. Alpha= 0.1 Confidence= 0.9 df= 6 MSE= 527.5556 Critical Value of F= 3.46330 Minimum Significant Difference= 49.357 Comparisons significant at the 0.1 level are indicated by '***'. Simultaneous Simultaneous Lower Difference Upper SIZE Confidence Between Confidence Comparison Limit Means Limit 1 - 2 -16.69 32.67 82.02 1 - 3 24.31 73.67 123.02 *** 2 - 1 -82.02 -32.67 16.69 2 - 3 -8.36 41.00 90.36 3 - 1 -123.02 -73.67 -24.31 *** 3 - 2 -90.36 -41.00 8.36 Dependent Variable: TIME Contrast DF Contrast SS Mean Square F Value Pr > F svl1 1 8140.1667 8140.1667 15.43 0.0077 T for H0: Pr > |T| Std Error of Parameter Estimate Parameter=0 Estimate svl 73.6666667 3.93 0.0077 18.7537651 svm+l 53.1666667 3.27 0.0170 16.2412370