Significance tests - DA
|Top Previous Next|
This tab presents the results of a number of standard tests in a Rich Text box.
Wilks' lambda is used to test the significance of the discriminant function as a whole. Generally, you will want a probability less than 0.05, as this indicates a rejection of the null hypothesis at the 5% level that the groups have the same mean discriminant function scores. If the groups have different scores, then the model is discriminating between the groups.
DF is the degrees of freedom for Wilks' lambda.
Prob is the probability that the groups have the same discriminant scores.
Pillai's trace is a second overall test for a significant ability to discriminate between the groups.
Bartlett's Test of Sphericity tests if the variables are uncorrelated. If this is the case then the correlation matrix will be an identity matrix. The null hypothesis is that the correlation matrix comes from a population in which the variables are noncollinear (i.e. an identity matrix) and that the non-zero correlations in the sample matrix are due to sampling error. For a successful discriminant analysis you need to be able to reject the null hypothesis.
DF is the degrees of freedom for the Bartlett's test.
Prob is the probability that the variables are all uncorrelated.
Press Ctrl-Alt-C, or Edit: Copy All, to copy the entire text of this output. To copy a selected portion, select the text you require, and press Ctrl-C on your keyboard, or Edit: Copy.