[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

From |
Steven Samuels <sjhsamuels@earthlink.net> |

To |
statalist@hsphsun2.harvard.edu |

Subject |
Re: st: robust cluster in GLM |

Date |
Tue, 16 Sep 2008 19:31:51 -0400 |

Leny, that's not exactly the information I asked for, but there is probably nothing wrong with your result. If you have at least 2 time points per case, then a line can be estimated within each. Therefore all 80 cases will be utilized in the analysis.

-Steve

On Sep 16, 2008, at 5:12 PM, Leny Mathew wrote:

Sorry about that Steve, SE of the coefficient. Time is in weeks and is used as continuous in the model. The average at T1 is 10 weeks, T2 is 20 weeks and T3 is 29.7 weeks. Leny On Tue, Sep 16, 2008 at 4:27 PM, Steven Samuels <sjhsamuels@earthlink.net> wrote:-

Leny, "standard error" of what? Please follow the Statalist FAQ (section

3.3) and "Say exactly what you typed and exactly what Stata typed (or did)

in response. N.B. exactly!" Are the time-points equally spaced? If so, and

you are estimating the slope, then the middle time point would be ignored.

-Steve

On Sep 16, 2008, at 3:46 PM, Leny Mathew wrote:

Dear Statalisters,

I'm using a GLM model with robust cluster

option to model longitudinal data across three time points. I have the

dependent variable on 80 cases at T1 and T2, but have only 77 at the

third time point. When I run the model the number of observations is

237, but it says that the standard error is adjusted for 80 clusters.

Why would this be the case? How does stata take the missing info for

the 3 people at T3 into consideration when it calculates the

parameters and SE?

Thanks in advance for any suggestions,

Leny

*

* For searches and help try:

* http://www.stata.com/help.cgi?search

* http://www.stata.com/support/statalist/faq

* http://www.ats.ucla.edu/stat/stata/

* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

* * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/

**References**:**st: robust cluster in GLM***From:*"Leny Mathew" <lenymathewc@gmail.com>

**Re: st: robust cluster in GLM***From:*Steven Samuels <sjhsamuels@earthlink.net>

**Re: st: robust cluster in GLM***From:*"Leny Mathew" <lenymathewc@gmail.com>

- Prev by Date:
**st: foreach problem** - Next by Date:
**RE: st: Comparing datasets** - Previous by thread:
**Re: st: robust cluster in GLM** - Next by thread:
**st: detectable alternative/effect size;** - Index(es):

© Copyright 1996–2014 StataCorp LP | Terms of use | Privacy | Contact us | What's new | Site index |