why multivariate test table shows non significant value whereas univariate shows significant value
Please make it clear which multivariate test you have conducted and what do you mean with "univariate"?
I guess your question could probably be related to "ANOVA vs pairwise t-tests" case where similar question can arise.
Actually, it is well known that a global ANOVA F test can detect a difference of means even in cases where no individual [unadjusted pairwise] t-test of any of the pairs of means will yield a significant result.
I you are aware of this well-known tread which have been discussed by many people (see for example the following posts), and in the case where you are still not satisfied with my answer, I would say, the reason for the situation you encountered could be the lack of some key assumptions needed in your data for doing the tests you have implemented .
The posts which I mentioned above:
https://stats.stackexchange.com/questions/16665/how-can-i-get-a-significant-overall-anova-but-no-significant-pairwise-difference
https://help.xlstat.com/6741-how-interpret-contradictory-results-between-anova-and
Hope this help.
Yours,
Hamid
In multivariate analysis, the joint effect of multiple variables may not collectively reach statistical significance, as observed in the non-significant value in your test table. However, in univariate analysis, individual variables may demonstrate significant effects on the outcome variable, indicating their independent influence. This discrepancy highlights the importance of considering both collective and individual variable effects.
Reference:
Hair Jr., J. F., Black, W. C., Babin, B. J., & Anderson, R. E. (2019).
Multivariate Data Analysis (8th ed.). Cengage.)
When performing multivariate analyses, you are taking into account the joint effect of several predictor variables. If some of those variables are correlated or some interaction between variables are considered, you may or may not observed statistical significance. Some variables may also mask the effect of other variables. This is normal and it does not mean that variables with non significance in your multivariate analysis are not important. It means that when they are considered with other variables, their effect may not be as important as others. Therefore, is important to explore both univariate and multivariate analyses in order to better know how your response variable is affected by your predictor variables.
- Badges
- Science topic
- Similar topics
- Chemistry
- Pharmacology
- Pharmaceuticals
- Drugs