[R] Handling survey data issue

Walter R. Paczkowski dataanalytics at earthlink.net
Thu Feb 2 15:44:18 CET 2006


Good morning,

I have a survey from a client that has a variable, say number of 
pills taken per day, that is coded as a numeric - in part.  The 
survey respondents were asked to state the number of pills and this 
number was recorded.  But, for those who took more than 10, the 
answer was recorded as 10+.  I have to calculate a mean and do other 
estimations. How do I handle the 10+?  Any suggestions on handling 
the 10+?  Is there a way to estimate a number for 10+?  I seem to 
recall a paper years ago that addressed the issue of a censored 
answer such as 10+ to find a number that could then be used in 
calculations.  This should be a common problem.  My next question, of 
course, is: Is there an R function for this?

Thanks for any help.

Walt Paczkowski

________________________

Walter R. Paczkowski, Ph.D.
Data Analytics Corp.
44 Hamilton Lane
Plainsboro, NJ  08536
(V) 609-936-8999
(F) 609-936-3733




More information about the R-help mailing list