Overview of this module Course 02429 Analysis of correlated data: Mixed Linear Models Module 7: The analysis of split-plot design data 1 Simple Split-plot designs Example: Tenderness of pork. 2 Split-plot with blocks Example: Split-plot with blocks, Yield of oats 3 Split-plot in perspective Per Bruun Brockhoff DTU Compute Building 324 - room 220 Technical University of Denmark 2800 Lyngby – Denmark e-mail: [email protected] Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 Fall 2014 1 / 16 Simple Split-plot designs b1 b3 b2 b4 a1 a1 a1 a1 Mixed Linear Models, Module 7 Simple Split-plot designs Basic structure a3 a3 a3 a3 Per Bruun Brockhoff ([email protected]) Fall 2014 2 / 16 Example: Tenderness of pork. Example: Tenderness of pork. b4 b3 b1 b2 a1 a1 a1 a1 b1 b3 b4 b2 a2 a2 a2 a2 b3 b2 b1 b4 a3 a3 a3 a3 b3 b1 b2 b4 a2 a2 a2 a2 b2 b1 b4 b3 Treatments are given on different levels 24 porks (pork) in 2 treatment groups (low pH/high pH).(pH) Each pork cut in half (right/left side). 2 cooling methods: One for each side. (C) 48 observations of tenderness. Factor structure: 24 2 [P]22 [I]48 22 Treatment effects are more easily found on finer levels For B: Randomized block setting with whole-plots as blocks. Mixed Linear Models, Module 7 Fall 2014 011 Ph × C41 For A: Completely randomized design with whole-plots as observational unit. Per Bruun Brockhoff ([email protected]) Ph1 4 / 16 Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 C21 Fall 2014 5 / 16 Simple Split-plot designs Example: Tenderness of pork. Simple Split-plot designs The split-plot mixed model for the example Example, test results Source of variation The model: Yi = α(pHi ) + β(Ci ) + γ(pH × Ci ) + d(Pi ) + εi phgroup cooling phgroup*cooling pH tested versus the pork (whole plot) variation: F = MSpH MSP Source of variation C and pH×C tested versus the residual error: F = Per Bruun Brockhoff ([email protected]) MSC×pH MSC , F = MSError MSError Mixed Linear Models, Module 7 Simple Split-plot designs phgroup cooling Fall 2014 6 / 16 Numerator degrees of freedom 1 1 1 Numerator degrees of freedom 1 1 Per Bruun Brockhoff ([email protected]) Example: Tenderness of pork. Denominator degrees of freedom 22 22 22 Denominator degrees of freedom 22 22 F P-value 8.67 2.25 0.18 0.0075 0.1479 0.6790 F P-value 8.67 2.33 0.0075 0.1403 Mixed Linear Models, Module 7 Split-plot with blocks Example, results summary Fall 2014 7 / 16 Example: Split-plot with blocks, Yield of oats Example: Split-plot with blocks, Yield of oats v3 v1 Variances: 2 σ̂W Example: Tenderness of pork. v2 = 1.2463, σ̂ 2 = 0.4725. v3 Fixed effects: v1 v2 α̂(low) = 5.6529, ([4.9240, 6.3819]) n3 n1 n0 n3 n0 n2 156 140 111 174 117 161 n2 n0 n1 n2 n1 n3 118 105 130 157 114 141 n2 n0 n0 n3 n1 n3 109 63 80 126 90 116 n3 n1 n2 n1 n2 n0 99 70 94 82 100 62 v3 n2 n1 n3 n1 n1 n2 104 89 122 89 103 132 n0 n3 n0 n2 n0 n3 70 117 74 81 64 133 n3 n2 n2 n0 n2 n1 96 89 112 68 132 129 n0 n1 n3 n1 n3 n0 60 102 86 64 124 89 v2 n1 n3 n3 n2 n0 n1 108 149 144 121 61 91 n2 n0 n1 n0 n3 n2 126 70 124 96 100 97 n2 n3 n3 n0 n0 n2 118 113 104 89 97 119 n0 n1 n2 n1 n1 n3 53 74 86 82 99 121 v1 v2 v1 v1 v3 α̂(high) = 7.1163, ([6.3873, 7.8452]) v2 v3 v1 Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 Fall 2014 8 / 16 Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 v2 v3 Fall 2014 10 / 16 Split-plot with blocks Example: Split-plot with blocks, Yield of oats Split-plot with blocks Example: Yield of oats Example: Split-plot with blocks, Yield of oats Yield of oats, exploration Average yield as a function of nitrogen level for each variety. Factors and their levels: v1 , v2 , v3 n0 , n1 , n2 , n3 1, 2, . . . , 18 1, 2, . . . , 6 V N P B Factor structure: [P]18 10 [I]72 45 V32 V × N12 6 Per Bruun Brockhoff ([email protected]) [B]65 011 N43 Mixed Linear Models, Module 7 Split-plot with blocks Fall 2014 11 / 16 Example: Split-plot with blocks, Yield of oats fertil variety fertil*variety Source of variation fertil variety Fall 2014 12 / 16 Split-plot in perspective Numerator degrees of freedom 3 2 6 Numerator degrees of freedom 3 2 Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 Split-plot in perspective Yield of oats, testing Source of variation Per Bruun Brockhoff ([email protected]) Denominator degrees of freedom 45 10 45 F 37.69 1.49 0.30 Denominator degrees of freedom 51 10 Mixed Linear Models, Module 7 P-value Treatments on different levels occur very often in practice More complicated and more than two levels may occur: <.0001 0.2724 0.9322 F Split-plot with correlated whole plots Whole plot conducted as an incomplete latin square A strip-split-split-plot design P-value The above are case studies in Littel et al. (1999). Serves as a kind of basis for repeated measures analysis. 41.05 1.49 <.0001 0.2724 Fall 2014 13 / 16 Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 Fall 2014 15 / 16 Split-plot in perspective Overview of this module 1 Simple Split-plot designs Example: Tenderness of pork. 2 Split-plot with blocks Example: Split-plot with blocks, Yield of oats 3 Split-plot in perspective Per Bruun Brockhoff ([email protected]) Mixed Linear Models, Module 7 Fall 2014 16 / 16
© Copyright 2026 Paperzz