statistics reporting

[teixeirak]

Here's the checklist and what we have at the moment:

[teixeirak]

Here's a running list of things to consider/do:

• [x] Do we need to report test statistics and everything else for the drought analyses we added on after the first round of comments? I was under the impression that these tests were only to get an idea of what effects were there, not to make any conclusions about a null hypothesis, so have not extracted them.
• [x] We don't have measures of variation for the quilt plot correlation (just Pearson's r and significance as threshold value) or the drought analyses (for the reason stated above). Moreover, we only report those r and p values in a table (SI) for 4 months. Verify with editor that this is okay.
• [ ] I'm not sure I should have checked the 2nd-to-last on hierarchical designs (not sure what "full reporting of outcomes" means).

[teixeirak]

As per an email from Anna Armstrong on 12/13, it seems we're okay with all this.