The POWER Procedure
- Overview
-
Getting Started
-
Syntax
PROC POWER Statement LOGISTIC Statement MULTREG Statement ONECORR Statement ONESAMPLEFREQ Statement ONESAMPLEMEANS Statement ONEWAYANOVA Statement PAIREDFREQ Statement PAIREDMEANS Statement PLOT Statement TWOSAMPLEFREQ Statement TWOSAMPLEMEANS Statement TWOSAMPLESURVIVAL Statement TWOSAMPLEWILCOXON Statement
-
Details
Overview of Power Concepts Summary of Analyses Specifying Value Lists in Analysis Statements Sample Size Adjustment Options Error and Information Output Displayed Output ODS Table Names Computational Resources Computational Methods and Formulas ODS Graphics ODS Styles Suitable for Use with PROC POWER
-
Examples
One-Way ANOVA The Sawtooth Power Function in Proportion Analyses Simple AB/BA Crossover Designs Noninferiority Test with Lognormal Data Multiple Regression and Correlation Comparing Two Survival Curves Confidence Interval Precision Customizing Plots Binary Logistic Regression with Independent Predictors Wilcoxon-Mann-Whitney Test
- References
| ODS Styles Suitable for Use with PROC POWER |
ODS styles control the appearance of graphs produced by PROC POWER. ODS provides over 50 styles, but most are not suitable for use in PROC POWER. PROC POWER requires a style that distinguishes curves based on a combination of color, line style, and symbol marker. Styles that are well-suited for use in PROC POWER include: STATISTICAL, ANALYSIS, DEFAULT, LISTING, and HTMLBLUECML. The HTMLBLUE and PLATEAU styles are commonly used, but they are not well-suited for use with PROC POWER because they rely primarily on color to distinguish curves rather than a combination of color, line style, and symbol marker.
In this chapter, a style is explicitly specified at the start of each example that uses ODS Graphics to remind you to use one of the suitable styles. Styles are specified in an ODS destination statement. Destinations include LISTING, HTML, RTF, PDF, and many others. You can set the style and the destination as follows:
ods html style=htmlbluecml;
ods graphics on;
proc power;
onesamplemeans
mean = 5 10
ntotal = 150
stddev = 30 50
power = .;
plot x=n min=100 max=200;
run;
ods graphics off;
ods html close;
For more information about ODS and ODS destinations, see Chapter 20, Using the Output Delivery System. For more information ODS styles, see Chapter 21, Statistical Graphics Using ODS.