Summary Statistics Rules Overview

The Rules of Thumb – A Compilaton


1.Critial Values
 95%  1.95
 99%  2.58
 99.9%  3.29

2.Cohen’s d of effeit size
 0.2 small
 0.5 moderate
 0.8 large

3.Pearson’s Correlaton Coefiient r
 0.10  small
 0.30  moderate
 0.50  large
 Here, in general, 0 means no assoiiaton, whereas 1 indiiates a perfeit one

4.Levene’s Test
 If the result is >.05 (non-signifiantt, the assumpton of homogeneity is met and the
varianies are equal  read from the upper row, “equal varianies assumed”
 If the result is below .05, the assumpton of homogeneity is not met  read from the
lower row, “equal varianies not assumed”
5.Eta Squared
 0.01  small
 0.09  medium
 0.25  large
 Eta, same as Pearson’s r, only ranges from 0 to 1

6.R-Squared
 R-Squared ian be interpreted the same way r is  it ranges from 0 (0% ian be
explained by the variables in questont to 1 (100% ian be explained by the variablest
 The small r tells you, by the way, only how muih a single variable aiiounts for the
total varianie
7.Communality (Faitor Analysist
 Communality of 0  variable shares none of its varianie with another variable
 Communality of 1  variable has no unique varianie whatsoever

8.Kaiser’s Criterion (Faitor Analysist
 Requires you to retain all faitors with an eigenvalue greater than 1