I'm mining social networks point data to use with GIS (Geographic information system). Obviously there are going to be more posts in areas with higher population which, if uncorrected, would end up resulting in a generalized population map. I'm trying to figure out how one may account for this bias. Would it be as simple as accounting for the population of that area? I have pretty fine-scale population counts (census block groups) which would allow for this. Could I take the inverse of the population (1/pop) so that areas with higher population have a lower weight (I'm planning on creating a kernel density map)?

  • 1
    $\begingroup$ What are you trying to get exactly? For most purposes, I'd think that this "bias" is actually a valuable part of the data. But if you want average posts per population, then multiplying by 1/pop seems fine to me. But you might want something else. $\endgroup$ – Peter Flom Sep 22 '13 at 15:50
  • $\begingroup$ I'm trying to get block-level case counts of illnesses to use as independent variables. Wouldn't the higher count of cases in, say, New York City be mainly a factor of the higher population in these areas and as such cause serious problems in the results? $\endgroup$ – Ross Wardrup Sep 22 '13 at 17:53
  • $\begingroup$ If you are doing it on a block level, then you would want to adjust on a block level; but whether and how you should adjust depends on your dependent variable and your whole set-up. If both DV and IV are population dependent, then you may not want to adjust $\endgroup$ – Peter Flom Sep 22 '13 at 20:25

Your Answer

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

Browse other questions tagged or ask your own question.