Final edits for publication.
Differences from Revision1:
- Use of 95th percentile instead of mean for sim dissimilarity scores. The 95th percentile and mean vary together, but the 95th percentile directly shows how dissimilar the observed SAD would have to be to count as "extreme" and is the most direct measure of the magnitude of the observed deviation.
- Updating appendix A1 to include new metrics in revision (I missed this during the revision)
- Figure edits to fit formatting requirements