Forecasting using SAS Forecast Server, SAS/ETS, and more

Has anyone attempted to use jittering to accomodate quantile regression with count data

Reply
Contributor
Posts: 46

Has anyone attempted to use jittering to accomodate quantile regression with count data

I  have count data and am interested in quantile regression. There has been some work in this area using ' jittering'  which adds a random uniform component to values creating a continuous dependent variable which can then be used in QR. I'm not sure how this impacts things like standard errors etc. There are packages in R and Stata that handle this, but I'm curious if anyone has attemtped this in SAS or has any suggestions. A couple references:

Reforming health care: Evidence from quantile
regressions for counts

Rainer Winkelmann

Journal of Health Economics 25 (2006) 131–145


"Basically, the approach transforms the discrete data problem into a
continuous data problem by adding a random uniform variable to each count. The
quantile regression functions of the transformed variable can then be estimated
using standard quantile regression software. To interpret the results, one can
compare the freely estimated quantile functions to those implied by the
respective Poisson or negative binomial estimates in order to detect excess
sensitivity in specific parts of the distribution, such as the lower or upper
tails."


See also:

Machado, J.A.F. and Santos Silva, J.M.C. (2005), Quantiles for Counts,
Journal of the American Statistical Association, vol. 100, no. 472, pp. 1226-1237.



SAS Super FREQ
Posts: 3,473

Re: Has anyone attempted to use jittering to accomodate quantile regression with count data

Would it be possible to do a quantile regression of the log(counts)?  Back before there were special procedures for regression of count data, that was a standard technique.  For example, here's a log-linear analysis of (the mean of) count data by using PROC GENMOD: SAS/STAT(R) 13.1 User's Guide

Ask a Question
Discussion stats
  • 1 reply
  • 1086 views
  • 0 likes
  • 2 in conversation