Re: st: Reconstructing the distribution from interval data

[Richard Goldstein <richgold@ix.netcom.com>]

Ronan,

I don't think there are "good" solutions but I would start with some of
Daniel Heitjan's work (e.g., Heitjan, D (1989), "Inference from grouped
continuous data: a review," _Statistical Science_, 4, #2: 164-183

another option would be to treat this as a missing data problem; for
multiple imputation is Stata you would need to use -ice- rather than the
official programs since -ice- allows you to put constraints (bounds) on
the imputed values

Rich

On 9/8/10 12:25 PM, Ronan Conroy wrote:
> I'm reviewing a paper in which doctors were asked how many times they
> had carried out a particular procedure in the last year. The
> questionnaire used categories, so responses looked like this:
> 
> Number of  Gynaecologists Plastic
> times                     Surgeons
> 
> None           17           3
> 1-5            24           34
> 5-10            0           4
> 10-20           0           1
> More than 20    0           1
> 
> I am wondering how one might calculate the likeliest underlying annual
> rate for each group based on the data above.
> 
> Has anyone come up against this problem and found a solution? I'd be
> very curious about the answer for practical reasons - the authors have
> data that are pretty hard to come by, and being able to calculate an
> annual rate would be very helpful indeed for future research.
> 
> Ronán Conroy
> Associate Professor
> Division of Population Health Sciences
> =================================
> 
> rconroy@rcsi.ie
> Royal College of Surgeons in Ireland
> Epidemiology Department,
> Beaux Lane House, Dublin 2, Ireland
> +353 (0)1 402 2431
> +353 (0)87 799 97 95
> +353 (0)1 402 2764 (Fax - remember them?)
> http://rcsi.academia.edu/RonanConroy
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/