Hello together,
I have run a multigroup comparison as proposed by Chin and as discussed in this forum, but still have one question regarding my results:
Using the complete sample, all path coefficients of my model are significant.
The result of my multigroup comparison, using 2 split samples (e.g. companies with high revenue and companies with low revenue, devided at the median of the complete sample), is that both subsamples lead to no significant paths.
How is that possible?
How can I interprete the result?
Thanks and regards
Sandro