I'm doing a project that's looking at a certain disease outbreak and vaccination in California counties using arcmap. I have aggregated data joins for disease outbreak rate, personal belief exemptions, vaccine coverage, and other demographic data (ethnicity, income...). All of the data layers are joined to the CA counties, so location differences is not an issue.

I am looking for a way to analyze the spatial distributions of, for example, personal belief exemptions and a certain ethnicity, to see if the data clusters in similar locations. I was thinking about spatial autocorrelation, but I am not sure if that applies for more than one variable.

  • Are you looking to see how multiple variables affect another, or simply want to see the spatial distribution of variables? And at what scale are you looking (i.e. local vs. global)? – Barbarossa Jul 8 '15 at 18:47
  • I wanted a way to compare the spatial distributions, and the scale is local. – Naomi Gruber Jul 9 '15 at 13:05

Spatial autocorrelation does look at whether two phenomena have similar spatial distribution, however I believe what you want goes a bit beyond that. You're looking at a regression analysis, which explores correlation between several independent or explanatory variables and an occurrence (dependent variable). In your case, why is this disease outbreak concentrated where it is - because of vaccinations, income levels as indications of access to health care, ethnicity, etc.

Some further resources on regression analysis:

| improve this answer | |
  • 1
    Unfortunately, in this case, the world according to ESRI is quite inadequate and distinctly NOT spatial models. One recommendation would be application of something in the family of scan statistics that allows for specification of multivariate and multitemporal processes models. Alternately, stepping into regression models one can apply a spatial autoregression (SAR) or conditional autoregression (CAR) approaches. Autoregressive models allow for specification of different hypothesis of spatial relationships and well as definition of explicit spatial process (eg., diffusion) terms. – Jeffrey Evans Jul 8 '15 at 21:01

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.