Download Analysis of a Split-Plot Experiment using SAS: SAS Program and Output - Prof. Mervyn G. Ma and more Study notes Statistics in PDF only on Docsity!
Example^ F17 SAS Programdata^ paper;do^ temp=^200 ,^225 ,^250 ,^275 ; do^ rep=^1 to^^3 ; do method=^1 to^^3 ; input^ strength @;output;end;end;end;datalines;30 34 29 28 31 31 31 35 3235 41 26 32 36 30 37 40 3437 38 33 40 42 32 41 39 3936 42 36 41 40 40 40 44 45; run ; proc^ glm^ data=paper;class^ rep method temp;model^ strength = rep method repmethod temp methodtemp;test^ h=method^ e=repmethod;lsmeans^ methodtemp/slice=method;lsmeans^ method/pdiff^ cl^ adj=tukey
e=rep*method;
contrast^ 'Temp:Linear Trend'^ temp -
3 -^1 1 3 ;
contrast^ 'T1 vs T2 @ M1'^ temp^^1
-^1 0 0 method*temp^^1 -^1 0 0 0 0 0 0 0
contrast^ 'T1 vs T3 @ M1'^ temp^^1
0 -^1 0 method*temp^^1 0 -^1 0 0 0 0 0 0
contrast^ 'T1 vs T4 @ M1'^ temp^^1
0 0 -^1 method*temp^^1 0 0 -^1 0 0 0 0 0
title^ 'Analysis of a Split-Plot Experiment using PROC GLM'; run ; --------------------------------------------------------------------------------------
proc^ sort^ data=paper;by^ temp method; run ; proc^ means^ data=paper^ noprint^ mean;by^ temp method;var^ strength;output^ out=meandat^ mean=cellmean; run ; title1^ c=firebrick h=^2 'Analysis of Strength of Paper';title2^ c=cornflowerblue h= 1.
f=centx^ 'Profile Plot of Cell Means';symbol1 c=crimson i=join v=square^ h= 1.5^ l=^2 ; symbol2 c=darkorange i=join v=diamond^ h= 1.5^ l=^3 ; symbol3 c=cadetblue i=join v=triangle^ h= 1.5^ l=^4 ; axis1 c=magenta label=(c=dodgerblue^ h=^1 f=swissu^ a=^90 'Cell Means')value=(c=purple);axis2 offset=( .2 in) label=(c=dodgerblue^ h=^1 f=swissu^ 'Levels of Temperature')value=(c=purple); proc gplot data=meandat;plot cellmean*temp=method/vaxis=axis1^ haxis=axis2^ hm=^4 ; run ;
Analysis^ of^ a Split-Plot Experiment^
using^ PROC^ GLM^2 The GLM Procedure Dependent Variable: strength
Sum of Source^ DF^
Squares^ Mean^ Square^ F^ Value^ Pr >
F
Model^17 751.
44.2042484^ 11.13^ <.
Error^18 71.
Corrected^ Total^35 822.9722222R-Square^ Coeff^ Var^
Root MSE^ strength^ Mean0.913120 5.531963 1.993043^ 36. Source^ DF^ Type
I SS^ Mean^ Square^ F^ Value^ Pr >
F
rep^2 77.
38.7777778^ 9.76^ 0.
method^2 128.
64.1944444^ 16.16^ <.
rep*method^4 36.
9.0694444^ 2.28^ 0.
temp^3 434.
144.6944444^ 36.43^ <.
method*temp^6 75.
12.5277778^ 3.15^ 0.
Source^ DF^ Type III SS
Mean^ Square^ F^ Value^ Pr >^ F rep^2 77.
38.7777778^ 9.76^ 0.
method^2 128.
64.1944444^ 16.16^ <.
rep*method^4 36.
9.0694444^ 2.28^ 0.
temp^3 434.
144.6944444^ 36.43^ <.
method*temp^6 75.
12.5277778^ 3.15^ 0.
Tests of^ Hypotheses Using the Type^ III
MS^ for rep*method as an Error^ Term Source^ DF^ Type III SS
Mean^ Square^ F^ Value^ Pr >^ F method^2 128.
64.1944444^ 7.08^ 0.0485 Analysis of a Split-Plot Experiment^ using^ PROC^ GLM^3 The GLM ProcedureLeast Squares Meansstrengthmethod temp LSMEAN 1 200 29.6666667 1 225 34.6666667 1 250 39.3333333 1 275 39.0000000 2 200 33.3333333 2 225 39.0000000 2 250 39.6666667 2 275 42.0000000 3 200 30.6666667 3 225 30.0000000 3 250 34.6666667 3 275 40.
Analysis^ of^ a Split-Plot Experiment^
using^ PROC^ GLM^5 The GLM ProcedureLeast Squares MeansAdjustment for Multiple Comparisons: Tukey Standard Errors and^ Probabilities^ Calculated Using
the Type^ III^ MS^ forrep*method as an Error Termstrength LSMEANmethod LSMEAN Number 1 35.6666667 (^12) 38.5000000 (^23) 33.9166667 (^3) Least Squares Means for effect methodPr > |t| for H0: LSMean(i)=LSMean(j)Dependent Variable: strength i/j^1
1 0.^
2 0.^
3 0.4129^ 0.0434 strengthmethod LSMEAN^ 95% Confidence
Limits 1 35.666667^ 33.
2 38.500000^ 36.
3 33.916667^ 31.
Least^ Squares^ Means^ for^ Effect^ methodDifference^ Simultaneous
95%Between Confidence Limits for i^ j^ Means^ LSMean(i)-LSMean(j) 1 2 -2.833333^ -7.
1 3 1.750000^ -2.
2 3 4.583333^ 0.
Analysis^ of^ a Split-Plot Experiment^
using^ PROC^ GLM^6 The GLM Procedure Dependent Variable: strengthContrast^ DF^ Contrast SS
Mean^ Square^ F^ Value^ Pr >^ F Temp:Linear Trend^1 432.
432.4500000^ 108.87^ <.
T1^ vs^ T2 @^ M1^1 37.
37.5000000^ 9.44^ 0.
T1^ vs^ T3 @^ M1^1 140.
140.1666667^ 35.29^ <.
T1^ vs^ T4 @^ M1^1 130.
130.6666667^ 32.90^ <.
