We have a number of graphs which detail bat activity (in various ecoregions/sites) against temperature and wind speeds. We would like to run some sort of statistical analysis to determine if there is a relationship between the different ecoregions? With the naked eye, we can see there are certain outliers and other graphs almost overlapping but need to validate this with some statistical method. Thanks in advance.
Dear Joshua
Pl see the papers. May be of some help.
Kind REgards
Rajiv
Deleted research item The research item mentioned here has been deleted
Article Population dynamics of P.reflexa and D.erandantis defoliator...
Dear Joshua,
If you are looking for a linear relationship then you could try a general linear model. Bat activity will be your dependent variable and temperature and wind speed will be your covariates. The ecoregions will be a nominal factor - you may need to transform this into a number of "dummy" binary variables. This model has assumptions that you should first check. However, I would not recommend removing outliers just because they affect the validity of a model. I would only recommend removing them if the validity of the readings themselves is in question.
- Badges
- Science topic
- Similar topics
- Medicine
- Physiology
- Biosignals
- Biomedical Signal Processing
- Statistical Analysis