Comparing Treatment Means: Single Factor Experiment 


 The following lines from (title ... to RUN;) would go in a file, say
gpici.sas. This is the file you would run by typing 

 sas gpici.sas

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

title 'Single Factor Experiment: Wound Healing';

options linesize=79;

data wounds;
   infile 'gpi.dat';

input Material Gpi;

PROC ANOVA;
    CLASS Material;
    MODEL Gpi = Material;
    MEANS Material / lsd bon tukey duncan alpha = 0.05;

RUN;

%%%%%%%%%%%%%%%%%%%%%%%%%%
 
 Under the MEANS option: lsd = least significant difference
                         bon = Bonferroni
			 tukey = Tukey's studentized range
			 duncan = Duncan's multiple range
			 alpha = alpha (of course!)



%%%%%%%%%%%%%%%%%%%%%%%%%%%%

                    Single Factor Experiment: Wound Healing                   1


                              The ANOVA Procedure
 
                             t Tests (LSD) for Gpi

    NOTE: This test controls the Type I comparisonwise error rate, not the 
                          experimentwise error rate.


                     Alpha                            0.05
                     Error Degrees of Freedom           12
                     Error Mean Square            19.44444
                     Critical Value of t           2.17881
                     Least Significant Difference   7.8446


          Means with the same letter are not significantly different.
 
 
             t Grouping          Mean      N    Material

                      A        73.333      3    3       
                                                        
                      B        65.000      3    1       
                      B                                 
                      B        63.333      3    2       
                                                        
                      C        23.333      3    4       
                                                        
                      D        11.667      3    6       
                      D                                 
                      D         6.667      3    5       
                    Single Factor Experiment: Wound Healing                   4

                              The ANOVA Procedure
 
                     Duncan's Multiple Range Test for Gpi

    NOTE: This test controls the Type I comparisonwise error rate, not the 
                          experimentwise error rate.


                       Alpha                        0.05
                       Error Degrees of Freedom       12
                       Error Mean Square        19.44444


    Number of Means          2          3          4          5          6
    Critical Range       7.845      8.211      8.433      8.580      8.682


          Means with the same letter are not significantly different.
 
 
        Duncan Grouping          Mean      N    Material

                      A        73.333      3    3       
                                                        
                      B        65.000      3    1       
                      B                                 
                      B        63.333      3    2       
                                                        
                      C        23.333      3    4       
                                                        
                      D        11.667      3    6       
                      D                                 
                      D         6.667      3    5       

                 Tukey's Studentized Range (HSD) Test for Gpi

    NOTE: This test controls the Type I experimentwise error rate, but it 
             generally has a higher Type II error rate than REGWQ.


                 Alpha                                   0.05
                 Error Degrees of Freedom                  12
                 Error Mean Square                   19.44444
                 Critical Value of Studentized Range  4.74775
                 Minimum Significant Difference        12.087


          Means with the same letter are not significantly different.
 
 
            Tukey Grouping          Mean      N    Material

                         A        73.333      3    3       
                         A                                 
                         A        65.000      3    1       
                         A                                 
                         A        63.333      3    2       
                                                           
                         B        23.333      3    4       
                         B                                 
                    C    B        11.667      3    6       
                    C                                      
                    C              6.667      3    5       

                       Bonferroni (Dunn) t Tests for Gpi

    NOTE: This test controls the Type I experimentwise error rate, but it 
             generally has a higher Type II error rate than REGWQ.


                    Alpha                              0.05
                    Error Degrees of Freedom             12
                    Error Mean Square              19.44444
                    Critical Value of t             3.64889
                    Minimum Significant Difference   13.138

          Means with the same letter are not significantly different.
 
              Bon Grouping          Mean      N    Material

                         A        73.333      3    3       
                         A                                 
                         A        65.000      3    1       
                         A                                 
                         A        63.333      3    2       
                                                           
                         B        23.333      3    4       
                         B                                 
                    C    B        11.667      3    6       
                    C                                      
                    C              6.667      3    5