Models for Overdispersed Censored Counts
I am constructing models of counts of households in various income
categories. The income categories begin with $0 to $4999, continue with
$5,000 to $9,999, etc. The last income category is right censored,
aggregating all households with incomes greater than
$150,000. The households/ counts are overdispersed. I have
experimented with quasi-log-mu standard glm and with glm.nb from
Venables & Ripley/s MASS library. The preliminary results are very
fruitful but insufficient because the quasi-log-mu glm and glm.nb do
not account for the right censorship (and possibly may not account for
the unobserved heterogeneity). A recent article by Gurmu, Rilstone, &
Stern, in the //Journal of Econometrics// 88(1999):123-150, provides
new methods for semiparametric estimation of count regression models.
Their codes, however, in GAUSS, and I have been, heretofore, a loyal
SPlus user. Therefore, I am searching for Splus ways (especially with
glm, but perhaps in other ways too) of analyzing overdispersed censored
counts. Please let me know.
Sincerely,
Michael
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
-----------------------------------------------------------------------
This message was distributed by s-news@wubios.wustl.edu. To unsubscribe
send e-mail to s-news-request@wubios.wustl.edu with the BODY of the
message: unsubscribe s-news
|