[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]
Re: st: OLS regression versus survival analysis
THe fundamental problem, as I understand it, is with the distribution of the time to event. In most situations, after you fit an OLS model, the residuals will not be normally distributed. Hypothesis testing and coinfidence intervals will thus not be valid. This is the rationale
for using survival models.
ellen mastenbroek wrote:
> Hi Ronan,
> Thanks for your answer. However, I may not have been clear enough in stating my question: I meant OLS with time until the event as the dependent variable.
> As to your suggestion: I have already run the KM graphs, as well as Cox regression. But then I thought: why bother if right censoring is only 3 percent... (and there are no TVCs). Especially because in this case I think the OLS interpretation makes more sense theoretically...
> On Wed, 12 Jun 2002 15:10:40
> Ronan Conroy wrote:
> >on 12/6/02 1:05 PM, ellen mastenbroek at email@example.com wrote:
> >> I have set out to do a survival analysis, but it has become clear that only 3
> >> % of the cases are censored. Is it advisable to then carry out regular OLS
> >> regression instead (the advantage being that OLS results are more easy to
> >> interpret)?
> >No. Your endpoint is binary. You want to figure out what effect does a
> >particular predictor have on the risk of an event happening. That's a hazard
> >Start with Kaplan Meier graphs, which are even easier to interpret than OLS
> >regression, and work from there.
> >Ronan M Conroy (firstname.lastname@example.org)
> >Lecturer in Biostatistics
> >Royal College of Surgeons
> >Dublin 2, Ireland
> >+353 1 402 2431 (fax 2329)
> >And now, Mr President, how about the global alliance against climate change?
> >* For searches and help try:
> >* http://www.stata.com/support/faqs/res/findit.html
> >* http://www.stata.com/support/statalist/faq
> >* http://www.ats.ucla.edu/stat/stata/
> WIN a first class trip to Hawaii. Live like the King of Rock and Roll
> on the big Island. Enter Now!
> * For searches and help try:
> * http://www.stata.com/support/faqs/res/findit.html
> * http://www.stata.com/support/statalist/faq
> * http://www.ats.ucla.edu/stat/stata/
| Murray M Finkelstein PhD MD Phone: 416-326-7879
| Medical Consultant, FAX: 416-326-7761
| Ontario Ministry of Labour E-mail: email@example.com
| 655 Bay Street, 14th Floor
| Toronto, Ontario
| Canada M7A 1T7
* For searches and help try: