Logistic regression and rare events
- From: "Tue Lehn-Schiøler" <tuelehn@xxxxxxxxx>
- Date: Thu, 24 Jan 2008 06:01:03 -0800 (PST)
Previous threads in this group have discussed "Logistic regression and
rare events"
One of these threads mentions the reference:
G King and L Zeng (2000), "Logistic regression in rare events data,"
_Political Analysis_, 9: 1-27
In this paper it is stated (without reference) that:
"[for weighting] the usual method of computing standard errors is
severely biased"
I would appeciate if anyone can point me to a reference to this fact
or provide comments on the statement.
Other references dealing with the influence on rare events in logistic
regression are also welcome. Especially if they also deal with the
influence of rare events on confidence intervals and wald statistics.
/Tue
.
- Prev by Date: for vector q: P{(q,s) > 0 | s[i]>=0, sum(s[i])=1} =?
- Next by Date: Re: C-statistic
- Previous by thread: for vector q: P{(q,s) > 0 | s[i]>=0, sum(s[i])=1} =?
- Next by thread: JOB: Data Analyst - Loudoun County Public Schools
- Index(es):
Relevant Pages
|
