【正文】 s two factor interactions: General Linear Model Factor Type Levels Values Precoat fixed 2 Yes No Pressure fixed 5 2 4 5 7 9 Speed fixed 7 150 175 200 225 250 275 300 Analysis of Variance for Lubricat, using Adjusted SS for Tests Source DF Seq SS Adj SS Adj MS F P Precoat 1 Pressure 4 Speed 6 Precoat*Pressure 4 Precoat*Speed 6 Pressure*Speed 24 Error 24 Total 69 Unusual Observations for Lubricat Obs Lubricat Fit StDev Fit Residual St Resid 1 9 57 64 R denotes an observation with a large standardized residual. The effects of Precoat and Pressure are significant and there is a significant PrecoatPressure interaction. Speed even closer to being significant. Since there are no replicates error is 3 factor interactions 131 Minitab Follow Along: Analyzing the Design, cont. Minitab Commands Stat DOE Factorial Plots… Main Effects Setup... and Interaction Setup… (specify response and all factors) Precoat Pressure Die Speed 1 2 3 4 5 Lubrication Main Effects Plot Data Means for Lubrication 1 3 5 1 3 5 P r e c o a t P r e s s u r e D i e S p e e d Y e s N o 2 4 5 7 9 I n t e r a c t i o n P l o t D a t a M e a n s f o r L u b r i c a t i o n 132 Blocking Definition A block is a carefully chosen subset of the total runs in the experiment that are tested close together in time or under similar conditions. ? “Similar conditions” means that factors not being tested in the experiment are the same or very nearly so (includes factors such as staffing, physical environment, source materials, etc.) ? The experimental runs within a block are performed under the same conditions。 but different blocks encounter different conditions ? If a design is replicated, the replicates can be run in blocks instead of totally randomized ? If a design is not replicated, the trials can still be run in blocks。 the cost is increased confounding (the Minitab menu will select a blocked design for you with the least confounding possible) 133 Situations Where Experiments May Need to Be Blocked ? Can’t do all runs at o