Search results

  1. D

    quantile regression with a bounded dependent variable

    Quantile regression is cool. But I wonder about using it when the dependent variable is bounded. I often work with happiness variables, emerging from a questionnaire where respondents select from a 0 - 10 scale. Running a quantile regression, it seems that pretty much every independent...
  2. D

    Feedback on specific problem from "Understanding Society"?

    Wave 5 of "Understanding Society" (the UK general/household panel dataset) has just been released. I've been full of anticipation... But I'm now thinking that I'm not going to find something I expected to find. For Wave 5, the questionnaire was supposed to ask about citizenship (i.e...
  3. D

    European Social Survey -- income variable

    If you have worked with the ESS and are familiar with the vagaries of the income variable, perhaps you can help me identify a solution to a specific problem. Beginning in Round 4, the ESS rationalized the income variable so that all respondents were presented with deciles pertaining to the...