I try to analysis the best correlation between the biotic and environmental dataset. My question is: if I have one variable for the biota dataset and more variable for the environmental dataset, is possible I apply the BEST analysis?
Actually you could correlate all at once by calculating a correlation matrix. Alternatively, you could do a multiple regression of biota on the environmental variables. similarly, you could do a multivariate regression of the environmental variables on biota.. Notice that if the response is categorical the regression is logistic. I have tried to add a little detail to Victor's response. I wrote it this way because I'm not clear on what the DV(s?) is(are).
Yes you can use BEST with a single response variable. For a single response variable you use Euclidean distance to make the 'target' matrix, then search for the subset of (transformed) explanatory variables that gives the best match (with whatever the appropriate resemblance measure for those variables is). Other methods are available (as alluded to in previous answers) but each has its own set of assumptions.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis