Doing More with SAS/ASSIST Software |
You can use the Anova item on the
Data Analysis menu to perform an analysis of variance (ANOVA). You can construct
a simple analysis of variance, analysis of covariance (ANCOVA), multivariate
analysis of variance (MANOVA), and repeated measures ANOVA. You must have
SAS/STAT software licensed to complete this task.
In this section, you perform an analysis of variance
on the VENEER table to show the relationship between the brand of veneer and
the amount of veneer that was worn away during testing.
For additional information on performing an analysis
of variance, refer to the GLM and ANOVA procedures in the
SAS/STAT User's Guide.
- To display the Analysis of Variance window, follow this
selection path:
Tasks |
|
Data Analysis |
|
Anova |
|
Analysis of Variance |
Analysis of Variance Window
- Use the
Table button
to select the SASUSER.VENEER table.
- Use the Dependent button
to select WEAR as the dependent column.
The dependent column contains the measurements to be
analyzed. The Select Table Variables window displays
all the numeric columns in the VENEER table except any columns selected as
BY or classification columns. If you have more than one dependent column,
you can generate a separate ANOVA model for each dependent column or one multivariate
model for the set of dependent columns.
- Use the Classification
button to select BRAND as the classification column.
The classification columns identify analysis groups.
The variation in the dependent column values is analyzed within and across
the classification groups to determine whether or not the classification columns
are significant sources of variation. The Select Table Variables window displays all the columns in the VENEER table except
for any columns selected as BY or dependent columns.
Each classification column is treated as a main effect
in the model. You can modify the ANOVA model to add interactions, nested effects,
covariates, and random effects by using the Additional options item in the Analysis of Variance
window.
- Follow this selection path:
Run |
|
Submit |
The first page of the
analysis appears showing the class level information.
Analysis of Variance Output
- Use
the scroll bars or the FORWARD command or
function key to display the next page of the analysis.
Analysis of Variance Output (continued)
Refer
to the GLM and ANOVA procedures in the
SAS/STAT User's Guide
for information about interpreting the statistics in this report.
- Return to SAS/ASSIST software from the Output window. See
Returning to SAS/ASSIST Windows from the Output Window
for more information.
|
Performing a Repeated Measures ANOVA |
Repeated measures designs are characterized by recording
several measurements over time or space on the same experimental unit. This
section shows how to perform a repeated measures ANOVA with SAS/ASSIST software.
The first part of the process is data entry. For this particular example,
you apply a logarithmic function to the data and then you perform the repeated
measures ANOVA.
This example uses the data from Example 7 in "The
GLM Procedure" chapter of
SAS/STAT User's Guide.
In this section, you enter the
raw data using
the Create Data task. For more information on this
task, refer to the "Entering Data Interactively" chapter in
Getting Started with the SAS System Using SAS/ASSIST Software.
- Follow this selection
path:
Tasks |
|
Data Management |
|
Create data |
|
Interactively |
|
Enter data in tabular form |
The Select a New SAS Table to Create
window appears.
- In the Table field,
type
GLMEX7
. If desired,
select Permanent, select a library in which to
store the table, and select OK.. Select Goback. The Define a New SAS Table
window appears.
- Define the columns in the new table by editing
this window so that it looks like the following display.
Define a New SAS Table Window With Columns Defined
You might need to
resize the window to see all six rows.
- Follow this selection path:
File |
|
Close |
If prompted,
select Yes to save the changes. The FSEDIT
window appears.
- Enter the following data into the table. Note
that the hist5 column has a missing value for row
6 (drug=Morphine, depl=N).
Data for Repeated Measures ANOVA Example
Row (Obs) |
drug |
depl |
hist0 |
hist1 |
hist3 |
hist5 |
1 |
Morphine |
N |
0.04 |
0.20 |
0.10 |
0.08 |
2 |
Morphine |
N |
0.02 |
0.06 |
0.02 |
0.02 |
3 |
Morphine |
N |
0.07 |
1.40 |
0.48 |
0.24 |
4 |
Morphine |
N |
0.17 |
0.57 |
0.35 |
0.24 |
5 |
Morphine |
Y |
0.10 |
0.09 |
0.13 |
0.14 |
6 |
Morphine |
Y |
0.12 |
0.11 |
0.10 |
|
7 |
Morphine |
Y |
0.07 |
0.07 |
0.06 |
0.07 |
8 |
Morphine |
Y |
0.05 |
0.07 |
0.06 |
0.07 |
9 |
Trimethaphan |
N |
0.03 |
0.62 |
0.31 |
0.22 |
10 |
Trimethaphan |
N |
0.03 |
1.05 |
0.73 |
0.60 |
11 |
Trimethaphan |
N |
0.07 |
0.83 |
1.07 |
0.80 |
12 |
Trimethaphan |
N |
0.09 |
3.13 |
2.06 |
1.23 |
13 |
Trimethaphan |
Y |
0.10 |
0.09 |
0.09 |
0.08 |
14 |
Trimethaphan |
Y |
0.08 |
0.09 |
0.09 |
0.10 |
15 |
Trimethaphan |
Y |
0.13 |
0.10 |
0.12 |
0.12 |
16 |
Trimethaphan |
Y |
0.06 |
0.05 |
0.05 |
0.05 |
- Follow this selection path to exit the
FSEDIT window:
File |
|
Close |
If prompted, select Yes
to save the changes.
- In order to minimize correlation between the mean
and the variance of the data, the logarithm of the data needs to be calculated.
Follow this selection path:
Tasks |
|
Data management |
|
Subset/Copy |
The Subset or Copy a Table
window appears.
- If the active table is not GLMEX7, use the Table button to select it from the WORK library, or, if you
chose to save the table permanently, the permanent library where you saved
the table.
- Use the Output Table
button. In the Output Table or View window, specify
GLM7OUT as the name of the output table. Select Temporary
or Permanent, as desired, before selecting Goback.
- Select Define new columns.
The Define or Modify a Column window appears.
- In the Column field,
type
LHIST0
.
- Select Initialize.
The Enter Numeric Expression window appears. Select Function, then Mathematical Functions.
The Select Data window appears.
- Select LOG(n). The Specify Arguments to a
Function window appears.
- In the Value for parameter
field, type
HIST0
and select OK,
- Select OK twice. The Define New Columns window
appears.
- Select Add new column.
Repeat steps 11 through 16 to create the LHIST1, LHIST3,
and LHIST5 columns, using the HIST1, HIST3, and HIST5 columns, respectively,
as parameters. When finished, select OK (instead
of Add New Column) from the Define
New Column to return to the Subset or Copy a Table window.
- Select Submit from
the Run menu. If desired, select OK, then Goback to view the table;
otherwise, select Cancel, then Goback.
- You are now ready to perform the repeated measures
ANOVA. Follow this selection path:
Tasks |
|
Data Analysis |
|
Anova |
|
Analysis of Variance |
The
Analysis of Variance
window appears.
- If the active table is not GLM7OUT, use the Table button to select the GLM7OUT
table.
- Use the Dependent button
to select the LHIST0, LHIST1, LHIST3, and LHIST5 columns.
- Use the Classification
button to select the DRUG and DEPL columns.
- Follow this selection path:
Additional options |
|
Model
effects |
|
Interactions |
The Interactions window
appears.
- Select DRUG, then *, then DEPL to construct the
DRUG*DEPL interaction. Select OK,
then Goback.
- Select Analysis type,
then Repeated measures, then Factor
names and levels. The Repeated measures Factors window appears.
- In this window, you specify the repeated factor
and the number of levels associated with that factor. In the first row, type
TIME
for Factor
Name and
4
for Number of Levels. Select OK.
- Select Factor Values.
The Repeated measures Factor Levels window appears.
- In this window, you specify the intervals for
the repeated factor. In the four active spaces under Level Values, type
0
,
1
,
3
, and
5
. Select OK.
- Select Factor Transformations.
The Repeated Measures Transformations window appears.
In this window, you specify a single-degree-of-freedom contrast. For descriptions
of each transformation, select Help.
- Select Polynomial,
then OK.
- Select Options. The Repeated Measures Options window
appears.
- Select Test within-subject effects to produce an analysis-of-variance table for each contrast
defined by the within-subject factors. Select OK.
- Select Goback three
times to return to the Analysis of Variance window.
- Select Submit from
the Run menu. The results of the analysis appear
in the Output window.
- If desired, compare these results to those in
Example 7 in "The GLM Procedure" chapter of
SAS/STAT User's Guide.
Copyright 1999 by SAS Institute Inc., Cary, NC, USA. All rights reserved.