Wenshui, I was in a similar situation as you are now years ago. SAS is not
really handy on this. My take on the difference between MSM and Markov
Transition model lies on the transition matrix where for MSM, the transition
matrix has no memory so to simplify everything.
On the other hand, IMHO, you may consider a Bayesian approach to this so
that it is actually easier to code in SAS. A Gibbs sampler may simplify the
estimation to a one conditional model in each iteration.
Wenshui, I wish I had time to tackle this problem, but there are just too
many things piled up in front of me.
On Sat, 8 Jan 2011 19:14:48 -0500, Wensui Liu <liuwensui@GMAIL.COM> wrote:
>what i asked was markov transition model instead of markov switching
>model. also, the model is not related to time series. instead, it is
>related to longitudinal data.
>On Sat, Jan 8, 2011 at 3:53 PM, Olivier Van Parys
>> Hey Kevin.
>> Appoligies. What I meant off course is "one of the other gurus".
>> Happy new year to all the sas listers - gurus and mere mortals like myself
>> I still think the Markov switching model question from Weimu is still of
>> interest. I just hope the time series add on is not required to implement
>> On Jan 8, 2011 5:50 PM, "Viel, Kevin" <firstname.lastname@example.org> wrote:
>>>> -----Original Message-----
>>>> From: SAS(r) Discussion [mailto:SAS-L@LISTSERV.UGA.EDU] On Behalf Of
>>>> Olivier Van Parys
>>>> Sent: Saturday, January 08, 2011 12:44 PM
>>>> To: SAS-L@LISTSERV.UGA.EDU
>>>> Subject: Re: markov transition model with binary response
>>>> Hi Wensui - this is a very interesting question. I was wondering if any
>>>> the Guru had gotten back to you on this?
>>> Yikes! Olivier Wensui *is* one of the Gurus :)
>>> Kevin Viel, PhD
>>> Senior Research Statistician
>>> Patient Safety & Quality
>>> International College of Robotic Surgery
>>> Saint Joseph's Translational Research Institute
>>> Saint Joseph's Hospital
>>> 5671 Peachtree Dunwoody Road, NE, Suite 330
>>> Atlanta, GA 30342
>>> (678) 843-6076: Direct Phone
>>> (678) 843-6153: Facsimile
>>> (404) 558-1364: Mobile
>>> Confidentiality Notice:
>>> This e-mail, including any attachments is the
>>> property of Catholic Health East and is intended
>>> for the sole use of the intended recipient(s).
>>> It may contain information that is privileged and
>>> confidential. Any unauthorized review, use,
>>> disclosure, or distribution is prohibited. If you are
>>> not the intended recipient, please delete this message, and
>>> reply to the sender regarding the error in a separate email.
>I keep saying that the sexy job
>in the next 10 years will be statisticians.
>-- HAL VARIAN