Hi! I usually run trend analysis on poverty rates between two two 5-year periods for all the ZCTAs in my county (currently 08-13, and 14-18). I am considering testing recent poverty rate estimates for all ZCTAs in my county against the overall poverty rate for the county to highlight ZCTAs with significantly higher/lower rates than the county overall. I am not a stats expert and wondering if anyone here could point out any statistical reason I shouldn't do it this way. Thanks!
Hi Mark,
I was doing some research on this question and came across your reply. Are you familiar with any resources about why it is appropriate to statistically compare a ZCTA in County X to County X overall? It was my understanding that since the ZCTA in question is part of County X, then the two groups you are comparing are not independent of each other. Any help would be greatly appreciated!