I have developed a special hearing test and carried out a study on the
effect of musical training on the test scores. Briefly, the results
are as follows: When subjects are categorized as "professionals",
"amateurs", and "non-musicians" (nominal variable), there is a large
difference between groups (unpaired t-tests). When a correlation is
calculated between the continuous variables "hours of previous musical
training" and "test scores", the result (as expected) is also highly
significant. However, this does not provide any information, whether
the effect is due to a priori differences between the groups (e.g.
musical talents; form of training) or to the time spent with training.
I therefore calculated correlations within each single group
(professionals, amateurs, non-musicians). In this case, in none of the
musician-groups a correlation "hours of training"-"test score" was
observed (despite a high variability in "hours of training" both in
amateurs and in professionals). This suggests that only between-group
differences are of importance(either because of differentces in
musical talents or the form of musical training) but that the time of
training, per se, does not have an effect.
However, before I accept this result I want to make sure that the loss
of correlation is not simply due to a reduction in sample size. The
total number of subjects was 62 (21 professionals, 22 amateurs, 19
non-musicians).
So it would be a great advantage to retain the sample size of 62,
while treating "musical experience" (the 3 nominal groups) as a
covariate (only preserving the effect of "hours of training"). My
question is whether it would be legitimate to use a covariate
("training group") that is (a) nominal and (b) directly related to a
variable of interest (professionals had a much higher number of
training hours than amateurs). Is there an appropriate statistical
procedure to solve this problem? By the way, I am using "Statview" for
statistical analyses.
Thank you very much for your support. With kind regards ...