|
|
Hi,
thanks a lot for the comments on this aspect. Lunney (1970 in J. of
Educational Measurement) argue that an ANOVA can be done if certain
conditions are met (basically sufficient degrees of freedom for error) by
using Monte Carlo techniques (thanks to William for a copy of this paper).
There are other tests, of course, which do not violate the assumptions as
an ANOVA does.
Kind regards,
Robinson
|
Locked
