I do not have the geostat toolbox, I need to interpolate the census tracts that had very unreliable data (calculated with CV > 40) in order to perform a proper OLS regression on the data I am looking at. I was wondering what the difference is between using areal interpolation in geostat toolbox on the census tract polygon or using IDW in spatial analyst toolbox on the tract's centroid.
My understanding is that you need to deal with missing data in order to perform a proper OLS regression,
If this is understood Regression analysis basics—Help | ArcGIS for Desktop
and this has been done http://www.arcgis.com/home/item.html?id=71a65d35688a4502b123cbdfc99afdee
then a pretty good background to proceed with this type of analysis would be appropriate. So as long as you have a game plan for dealing with missing data and how it might affect the appropriateness of your results that would be fine. Do note, the comment in there about getting a result that is completely wrong if you don't know what you are doing. Aren't statistics fun but at least there is failure in success.
Thanks for the resources! I have a BS in GIS and I have performed this type of regression in the classroom and as a research assistant. However, I did not have much experience with the missing data issue until now. yea-aah statistics with real world data!