1

would appreciate some help with the following problem.

While running a MANCOVA i received the following error message from SPSS: " Box's Test of quality of Covariance Matrices is not computed because there are fewer than two nonsingular cell covariance matrices". So I do not have the Box's M to check the homogeneity of variance ( using Box's at least).

Some initial trawling on the web indicates that the problem might be a DV i have that was created using the sum of three of my other DV's (i.e., overall trust was computed by summing trust A, trust B, and trust C). I dont want to exclude this variable as it is an important part of my RQ.

What are my options?

Mikhael
  • 11
  • 1

1 Answers1

0

Having the sum variable as an additional dependent variable adds no independent information to having the component parts. The multivariate tests are going to be the same with or without it in the model.

David Nichols
  • 840
  • 2
  • 9