Analysing 0 observations, whatever the filters on the Data page currently select.
Rank correlation, so a relationship counts even when it bends. Each cell uses only the rows where both variables are present, because coverage is very uneven: dropping rows missing any variable would compute the whole matrix on an unrepresentative remainder.
Strongest relationships
| Pair | ρ | n | Reading |
|---|
Correlation is not cause, and two confounds run through every row of this table. Season is the big one: high-elevation finds happen in summer and low ones in spring and autumn, which is why elevation and temperature appear to rise together, hold the month still and that relationship flattens to roughly zero. Effort is the other: people record where people go.