Effect Size
Some suggest the use of eigenvalues as effect size measures, however, this is generally not supported. Instead, the canonical correlation is the preferred measure of effect size. It is similar to the eigenvalue, but is the square root of the ratio of SSbetween and SStotal. It is the correlation between groups and the function. Another popular measure of effect size is the percent of variance for each function. This is calculated by: (λx/Σλi) X 100 where λx is the eigenvalue for the function and Σλi is the sum of all eigenvalues. This tells us how strong the prediction is for that particular function compared to the others. Percent correctly classified can also be analyzed as an effect size. The kappa value can describe this while correcting for chance agreement.
Read more about this topic: Discriminant Function Analysis
Famous quotes containing the words effect and/or size:
“As soon as I suspect a fine effect is being achieved by accident I lose interest. I am not interested ... in unskilled labor.... The scientific actor is an even worker. Any one may achieve on some rare occasion an outburst of genuine feeling, a gesture of imperishable beauty, a ringing accent of truth; but your scientific actor knows how he did it. He can repeat it again and again and again. He can be depended on.”
—Minnie Maddern Fiske (1865–1932)
“Men of genius are not quick judges of character. Deep thinking and high imagining blunt that trivial instinct by which you and I size people up.”
—Max Beerbohm (1872–1956)