Canpow: A Program that Estimates Effect or Sample Sizes Required for Canonical Correlation Analysis

Publisher Website
Google Scholar
Abstract
The importance of interpreting significance test statistics in the framework of sample' size and of evaluating power against Type II error has been increasingly recognized by researchers. The paper describes a computer program, written in BASIC source code, that can be employed to evaluate research situations in which canonical correlation analysis will be employed. The program, CANPOW, can be employed to evaluate the effects of sample size variations, given an actual or expected set effect size, and to evaluate the effects of variations in effect sizes, given an actual or expected sample size. To assist interpretation, comparisons are also presented in various graphs and figures.

This publication has 6 references indexed in Scilit: