Here are a few jointplots I created with seaborn (“percent_obese” is from Elizabeth’s original dataset, “OBESITY_CrudePrev” is from the PLACES 2021 Zip Code Tabulation Area-level obesity prevalence estimates):
Note: controlling for SES, altitude and race essentially eliminates this correlation (it becomes 0.026642, p=0.26, n=1764 counties, essentially no different from what you’d get by random chance.)
Just noticed this—thanks for fixing the plots!
Here are a few jointplots I created with seaborn (“percent_obese” is from Elizabeth’s original dataset, “OBESITY_CrudePrev” is from the PLACES 2021 Zip Code Tabulation Area-level obesity prevalence estimates):
Note: controlling for SES, altitude and race essentially eliminates this correlation (it becomes 0.026642, p=0.26, n=1764 counties, essentially no different from what you’d get by random chance.)
Just noticed this—thanks for fixing the plots!