Output from mpower2.sas

Source
0 Graphs

Two Way Manova Example: Power analysis 1

                  Plot of Y2*Y1.  Symbol is value of DRUG.

        |
        |
  12.00 +                                                                  C
  11.75 +
  11.50 +
W 11.25 +
e 11.00 +
e 10.75 +
k 10.50 +
2 10.25 +
  10.00 +
w  9.75 +
e  9.50 +
i  9.25 +
g  9.00 +
h  8.75 +            B
t  8.50 +                                                  C
   8.25 +         B A
l  8.00 +
o  7.75 +
s  7.50 +
s  7.25 +
   7.00 +
   6.75 +
   6.50 +
   6.25 +    A
        |
        --+------------+------------+------------+------------+------------+--
          6            8           10           12           14           16

                                   Week1 weight loss


Two Way Manova Example: Power analysis 2

                       General Linear Models Procedure
                           Class Level Information

                          Class    Levels    Values
                          SEX           2    F M

                          DRUG          3    A B C


                   Number of observations in data set = 24


Two Way Manova Example: Power analysis 3

                       General Linear Models Procedure
                      Multivariate Analysis of Variance

          Characteristic Roots and Vectors of: E Inverse * H, where
           H = Type I SS&CP Matrix for SEX   E = Error SS&CP Matrix

        Characteristic   Percent        Characteristic Vector  V'EV=1
             Root
                                                    Y1             Y2
            0.00751918    100.00            0.07175780     0.03444374
            0.00000000      0.00           -0.13423121     0.13423121


               Manova Test Criteria and Exact F Statistics for
                   the Hypothesis of no Overall SEX Effect
           H = Type I SS&CP Matrix for SEX   E = Error SS&CP Matrix

                             S=1    M=0    N=7.5

 Statistic                     Value          F      Num DF    Den DF  Pr > F
 Wilks' Lambda              0.99253694     0.0639         2        17  0.9383
 Pillai's Trace             0.00746306     0.0639         2        17  0.9383
 Hotelling-Lawley Trace     0.00751918     0.0639         2        17  0.9383
 Roy's Greatest Root        0.00751918     0.0639         2        17  0.9383
          Characteristic Roots and Vectors of: E Inverse * H, where
          H = Type I SS&CP Matrix for DRUG   E = Error SS&CP Matrix

        Characteristic   Percent        Characteristic Vector  V'EV=1
             Root
                                                    Y1             Y2
            4.57602675     98.63            0.14784109    -0.07693601
            0.06350991      1.37           -0.03619684     0.11526161


                Manova Test Criteria and F Approximations for
                   the Hypothesis of no Overall DRUG Effect
          H = Type I SS&CP Matrix for DRUG   E = Error SS&CP Matrix

Two Way Manova Example: Power analysis 4

                       General Linear Models Procedure
                      Multivariate Analysis of Variance

                            S=2    M=-0.5    N=7.5

 Statistic                     Value          F      Num DF    Den DF  Pr > F
 Wilks' Lambda              0.16862952    12.1991         4        34  0.0001
 Pillai's Trace             0.88037810     7.0769         4        36  0.0003
 Hotelling-Lawley Trace     4.63953666    18.5581         4        32  0.0001
 Roy's Greatest Root        4.57602675    41.1842         2        18  0.0001

         NOTE: F Statistic for Roy's Greatest Root is an upper bound.
                NOTE: F Statistic for Wilks' Lambda is exact.
          Characteristic Roots and Vectors of: E Inverse * H, where
        H = Type I SS&CP Matrix for SEX*DRUG   E = Error SS&CP Matrix

        Characteristic   Percent        Characteristic Vector  V'EV=1
             Root
                                                    Y1             Y2
            0.28372273     97.94           -0.00284433     0.09555092
            0.00596889      2.06           -0.15218117     0.10037136


                Manova Test Criteria and F Approximations for
                 the Hypothesis of no Overall SEX*DRUG Effect
        H = Type I SS&CP Matrix for SEX*DRUG   E = Error SS&CP Matrix

                            S=2    M=-0.5    N=7.5

 Statistic                     Value          F      Num DF    Den DF  Pr > F
 Wilks' Lambda              0.77436234     1.1593         4        34  0.3459
 Pillai's Trace             0.22694905     1.1520         4        36  0.3481
 Hotelling-Lawley Trace     0.28969161     1.1588         4        32  0.3473
 Roy's Greatest Root        0.28372273     2.5535         2        18  0.1056

         NOTE: F Statistic for Roy's Greatest Root is an upper bound.
                NOTE: F Statistic for Wilks' Lambda is exact.

Two Way Manova Example: Power analysis 5 OUTSTAT= data set

OBS  _NAME_  _SOURCE_  _TYPE_       Y1       Y2  DF       SS     F       PROB
 1     Y1    ERROR     ERROR    94.500   76.500  18   94.500    .       .
 2     Y2    ERROR     ERROR    76.500  114.000  18  114.000    .       .
 3     Y1    SEX       SS1       0.667    0.667   1    0.667   0.1270  0.72572
 4     Y2    SEX       SS1       0.667    0.667   1    0.667   0.1053  0.74934
 5     Y1    DRUG      SS1     301.000   97.500   2  301.000  28.6667  0.00000
 6     Y2    DRUG      SS1      97.500   36.333   2   36.333   2.8684  0.08292
 7     Y1    SEX*DRUG  SS1      14.333   21.333   2   14.333   1.3651  0.28056
 8     Y2    SEX*DRUG  SS1      21.333   32.333   2   32.333   2.5526  0.10569

Two Way Manova Example: Power analysis 6 Retrospective power analysis

                                        EFFECT      ALPHA
                     Power analysis for SEX SS1      0.05


                  ROOTS     THETA         S         M         N
              0.0075192 0.0074631         1         0       7.5
                      0         0

                   Value        F    df1   df2     Eta##2 Non-Cent.     Power

Wilks' Lambda      0.993    0.064      2    17     0.0075    0.0639    0.0541
Pillai's Trace     0.007    0.064      2    17     0.0075    0.0639    0.0541
Lawley Trace       0.008    0.064      2    17     0.0075    0.0639    0.0541
Roy's max. Root    0.008    0.064      2    17     0.0075    0.0639    0.0541


                                       EFFECT        ALPHA
                    Power analysis for DRUG SS1       0.05


                  ROOTS     THETA         S         M         N
              4.5760267 0.8206608         2      -0.5       7.5
              0.0635099 0.0597173

                   Value        F    df1   df2     Eta##2 Non-Cent.     Power

Wilks' Lambda      0.169   12.199      4    34     0.5894    24.398    0.9728
Pillai's Trace     0.880    7.077      4    36     0.4402    14.154    0.8174
Lawley Trace       4.640   18.558      4    32     0.6988    37.116    0.9983
Roy's max. Root    4.576   41.184      2    18     0.8207    82.368         1


                                   EFFECT                ALPHA
                Power analysis for SEX*DRUG SS1           0.05


                  ROOTS     THETA         S         M         N
              0.2837227 0.2210156         2      -0.5       7.5
              0.0059689 0.0059335

                   Value        F    df1   df2     Eta##2 Non-Cent.     Power

Wilks' Lambda      0.774    1.159      4    34       0.12    2.3187    0.1733
Pillai's Trace     0.227    1.152      4    36     0.1135     2.304    0.1734
Lawley Trace       0.290    1.159      4    32     0.1265    2.3175    0.1721
Roy's max. Root    0.284    2.554      2    18      0.221     5.107    0.4444