TWO WAY ANOVA WITH REPLICATION Also called a Factorial Experiment. Factorial Experiment is used to evaluate 2 or more factors simultaneously. Replication means an independent repeat of each factor combination. The purpose of factorial experiment is to examine: 1. The effect of factor A on the dependent variable, y. 2. The effect of factor B on the dependent variable, y along with 3. The effects of the interactions between different levels of the factors on the dependent variable, y. Interaction exists when the effect of a level for one factor depends on which level of the other factor is present. Advantages of Factorial Experiment over one factor at a time (one-way ANOVA) – more efficient & allow interactions to be detected. The effect model for a factorial experiment can be written as: i 1,...,a yijk i j ij ijk j 1,...,b k 1,...,r yijk : The response from the kth experimental unit receiving the ith level of factor A and the jth level of factor B : Overall mean i : An effect due to the ith level of factor A j : An effect due to the jth level of factor B ij : An interaction effect of the ith level of factor A with jth level of factor B ijk : A random error associated with the response from the kth experimental unit receiving the ith level of factor A combined with jth level of factor B There are three sets of hypothesis: 1. Factor A effect: 2. Factor B effect: 3. Interaction effect: The results obtained in this analysis are summarized in the following ANOVA table: Two way Factorial Treatment Structure B2 B1 A1 A2 A3 y111 y112 y11 . y11n y121 y122 y12 . y12 n y211 y212 y21 . y21n y221 y222 y22 . y22 n y311 y312 y31 . y31n y321 y322 y32 . y32n y1 y2 y1 y2 y3 y where SSA SSB yi br y ar 2 2 j y2 abr y2 abr 2 y ij y2 SSAB SSA SSB r abr 2 y SST yijk 2 abr SSE SST SSA SSB SSAB Example: The two-way table gives data for a 2x2 factorial experiment with two observations per factor – level combination. Factor B Factor A Level 1 2 1 29.6 35.2 47.3 42.1 2 12.9 17.6 28.4 22.7 Construct the ANOVA table for this experiment and do a complete analysis at a level of significance 0.05. Solution: Factor B Factor A Level 1 2 1 29.6 35.2 64.8 47.3 42.1 89.4 2 12.9 17.6 30.5 28.4 22.7 51.1 81.6 95.3 140.5 235.8 154.2 Solution: 1. Set up hypothesis Factor A effect: H 0 : 1 2 a 0 H1: at least one i 0 Factor B effect: H 0 : 1 2 b 0 H1: at least one j 0 Interaction effect: H 0 : ij 0 for all i, j H1: at least one ij 0 SST yijk 2 y2 abr 29.6 35.2 2 2 235.82 22.7 2 2 2 2 972.715 yi 2 y2 SSA br abr 154.22 81.62 4 658.845 235.82 8 y 2 2 y j SSB ar abr 95.32 140.52 235.82 4 8 255.38 2 y ij y2 SSAB SSA SSB r abr 64.82 89.4 2 30.52 51.12 235.82 658.845 255.38 2 8 2 SSE SST SSA SSB SSAB 972.715 658.845 255.38 2 56.49 2. Calculation (given the ANOVA table is as follows): Source of Variation SS df MS F A 658.845 1 658.845 46.652 B 255.38 1 255.38 18.083 2 1 2 0.1416 Error 56.49 4 14.1225 Total 972.715 7 AB 3. With = 0.05 we reject H 0 if : FA F ,a 1,ab r 1 for effect of factor A FB F ,b 1,ab r 1 for effect of factor B FAB F , a 1 b 1 ,ab r 1 for effect of interaction 4. From ANOVA/table F, the critical and F effects are given as follow: FA 46.652 and F ,a 1,ab r 1 F0.05 ,1,4 7.71 FB 18.083 and F ,b 1,ab r 1 F0.05 ,1,4 7.71 FAB 0.1416 and F , a 1b 1,ab r 1 F0.05 ,1,7 7.71 5. Factor A : since FA 46.652 >F0.05,1,4 7.71 , thus we reject H 0 We conclude that the difference level of A effect the response Factor B : since FB 18.083 > F0.05,1,4 7.71 , thus we reject H 0 We conclude that the difference level of B effect the response Interaction: since FAB 0.1416 F0.05,1,4 7.71 , thus we failed to reject H 0 We conclude that no interaction between factor A and factor B. Exercise: In a study to determine which are the important source of variation in an industrial process, 3 measurements are taken on yield for 3 operators chosen randomly and 4 batches a raw materials chosen randomly. It was decided that a significance test should be made at the 0.05 level of significance to determine if the variance components due to batches, operators, and interaction are significant. In addition, estimates of variance components are to be computed.The data are as follows, with the response being percent by weight. Batch Operator 1 2 3 4 1 66.9 68.1 67.2 68.3 67.4 67.7 69.0 69.8 67.5 69.3 70.9 71.4 2 66.3 65.4 65.8 68.1 66.9 67.6 69.7 68.8 69.2 69.4 69.6 70.0 3 65.6 66.3 65.2 66.0 66.9 67.3 67.1 66.2 67.4 67.9 68.4 68.7 Perform the analysis of variance of this experiment at level of significance 0.05. State your conclusion
© Copyright 2026 Paperzz