Logistic regression and proportional hazard model
Please, I have a problem to interpret (or to compare) the output of logistic regression and that of the Cox model. let : Y is a binary variable where Y=1 when event is present and X is a binary explatory variable, X=1 in the exposition case and 0 otherwise T is the time where the events are observed or censured. In logistic case (independently of time) we estimate: p(x)=Pr(Y=1/X=x) Odds(x)=Logit(p(x)) ( = Ln( P(x)/(1-p(x)) ) ) OR=Odds(1)/odds(0) = Exp(a) where a is the coefficient of X in logistic regression. The proportional hazard model, takes into account the time to estimate the instantaneous and relative risk: h(t,x)=ho(t)*Exp(bx) (hazard function) thus the ratio of estimated hazards for X=1 and X=0 is h(t,1)/h(t,0)=Exp(b) (presumedly independent of time) My problems: 1) In which case, Exp(a) (from logistic) estimates the relative risk (P(1)/P(0)) ? 2) In which case we expect that Exp(a) = Exp(b) (or approximate) ie the ratio of two odds estimates the ratio of tow hazards ie the logistic model and the Cox model give the same (or approximately) results ?? Thank you in advance. |===| | Hassane ABIDI (PhD) | | Unite d'Epidemiologie; Centre Hospitalier Lyon-Sud| | Pavillon 1.M, 69495 Pierre Benite Cedex, France | | Tel: (33) 04 78 86 56 87 ; Fax: (33) 04 78 86 33 31 | | E. mail: [EMAIL PROTECTED] | |===| = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Software
Of course you'll get what you pay for. "Comet" [EMAIL PROTECTED] wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... I search a good and free:) sofware of stat thanks = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Software (fwd)
- Forwarded message from Ken - Of course you'll get what you pay for. "Comet" [EMAIL PROTECTED] wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... I search a good and free:) sofware of stat - End of forwarded message from Ken - When I installed Win98 on my computer at home it crashed multiple times per day. I'm writing this on a FreeBSD system that supports thousands of users and crashes less than once a year. _ | |Robert W. Hayden | | Work: Department of Mathematics / |Plymouth State College MSC#29 | |Plymouth, New Hampshire 03264 USA | * |fax (603) 535-2943 /| Home: 82 River Street (use this in the summer) | )Ashland, NH 03217 L_/(603) 968-9914 (use this year-round) Map of New[EMAIL PROTECTED] (works year-round) Hampshire http://mathpc04.plymouth.edu (works year-round) The State of New Hampshire takes no responsibility for what this map looks like if you are not using a fixed-width font such as Courier. "Opportunity is missed by most people because it is dressed in overalls and looks like work." --Thomas Edison = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: stat question
David Heiser wrote: There is a lot of stat work involving maximum likelihood estimates, where there is no probability support unless you take a Bayesian approach. (Which is infrequent.) Cute = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Conference Announcement
+++ +++ CONFERENCE ANNOUNCEMENT +++ +++ Fifth Annual BEYOND THE FORMULA Introductory Statistics For A New Century: Integrating New Curriculum Ideas and Modern Techniques Into Our Beginning Statistics Course A Statistics Conference for Mathematics Teachers Teaching Introductory Statistics [This conference is for all teachers of Introductory Statistics, from those teaching the first time to the experienced teacher. There are sessions planned for all.] DATE: Thursday, August 2, 2001, 8:30 AM to 4:30 PM, and 6:30 PM to 9:00 PM Friday, August 3, 2001,8:30 AM to 3:00 PM LOCATION: Monroe Community College 1000 East Henrietta Road Rochester, NY 14623 AGENDA PLANS INCLUDES: Addresses by a keynote speaker, Two real world statistical case studies, Several sessions focusing on how to integrate new materials into the beginning course, Several sessions encompassing classroom teaching strategies and curriculum issues, Several hands-on computer (web and software) and calculator sessions, Dinner and an after-dinner speaker, Publisher's Book Exhibit And much more!! MAJOR THEMES INCLUDE: Teaching Techniques Technology Curriculum Applications It is clear from the discussions that occur on the various statistical lists that many members are doing interesting and innovative things in their statistics classrooms, and have much to offer other statistics instructors. Have you ever thought about making a presentation to share your innovations? You can find out more about the Beyond The Formula conferences by visiting our website at www.monroecc.edu/depts/math/beyond1.htm The Conference Steering Committee is interested in hearing from you about: presentation(s) you would like to make, topics you would like to see on our program, people you would be interested in hearing speak, and any other way that we could help you in your teaching of statistics. The Steering Committee's email address is: [EMAIL PROTECTED] = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: considering to study PhD
christopher ([EMAIL PROTECTED]) wrote: : Hi, : I am writing here in order to seek advice. I am working as a : statistical analyst. I'm a Hong Kong citizen and want to pursue a PhD. : Somebody told me that studying in UK take only 3 years since I have : already got a MPhil, is it the case? I ask this question because : studying in USA may take 5 years, it may be too long for me. Also, : what is the living standard in UK? How much pounds per month is needed : to survive? As I am doing a job now so I need to quit, will I find a : job easily after I have finished the PhD? I study PhD for fun, so I : don't want to take too much risk. One more thing, I want to study : Investment. : : Thanks The time needed to get a PhD in the US is not fixed. According to my experience an average student may need one year to prepare the prelims and another year to find advisor and a research subject and pass the oral exam. Then he or she may spend three more years to get the dissertation done (longer than that may make your advisor "impatient"). But if you are well prepared, have some research experience and manage to start working on your dissertation from the very first day, nothing prevents you from presenting it in three years if it is ready by then. You still have to take the prelims, the oral exam and whatever other formalities required by your university (including taking a number of graduate courses), but that could be done simultaneously to your dissertation work. Miguel A. Lerma = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
psychology and nominal data
Hi there, I am a student conducting an experiment about the McGurk Effect (where when a word is seen spoken while a different word is heard through headphones, the perceived word is an integration of the two). I am hoping to cue either the auditory word or the visual word before presenting the video and sound to see if that makes the word more detectable. My problem is that I will have to ask participants to select which word they hear from a choice of 3 - the auditory word, the visual word, and the McGurk integration. This therefore constitutes nominal data. The only statistical test I can find to use on nominal data is a chi square, is this right? My textbook says that I can only use a chi square if each participant only contributes one piece of data, and only participates in one condition, which is unrealistic in a study like this, so can I use chi sqaure, and if not, what should I use? Hope someone can help me, thanks, Kathryn = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Software (fwd)
Hi Bob, What is FreeBSD? Alan Bob Hayden wrote: - Forwarded message from Ken - Of course you'll get what you pay for. "Comet" [EMAIL PROTECTED] wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... I search a good and free:) sofware of stat - End of forwarded message from Ken - When I installed Win98 on my computer at home it crashed multiple times per day. I'm writing this on a FreeBSD system that supports thousands of users and crashes less than once a year. _ | |Robert W. Hayden | | Work: Department of Mathematics / |Plymouth State College MSC#29 | |Plymouth, New Hampshire 03264 USA | * |fax (603) 535-2943 /| Home: 82 River Street (use this in the summer) | )Ashland, NH 03217 L_/(603) 968-9914 (use this year-round) Map of New[EMAIL PROTECTED] (works year-round) Hampshire http://mathpc04.plymouth.edu (works year-round) The State of New Hampshire takes no responsibility for what this map looks like if you are not using a fixed-width font such as Courier. "Opportunity is missed by most people because it is dressed in overalls and looks like work." --Thomas Edison = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ = -- Alan McLean ([EMAIL PROTECTED]) Department of Econometrics and Business Statistics Monash University, Caulfield Campus, Melbourne Tel: +61 03 9903 2102Fax: +61 03 9903 2007 = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: psychology and nominal data
On Wed, 29 Nov 2000, Kathryn, alias [EMAIL PROTECTED], wrote: Hi there, I am a student conducting an experiment about the McGurk Effect (where when a word is seen spoken while a different word is heard through headphones, the perceived word is an integration of the two). I am hoping to cue either the auditory word or the visual word before presenting the video and sound to see if that makes the word more detectable. My problem is that I will have to ask participants to select which word they hear from a choice of 3 - the auditory word, the visual word, and the McGurk integration. This therefore constitutes nominal data. Non sequitur. The choice of words is nominal, indeed; whether your DATA are nominal is another matter entirely. Supposing your sample size is reasonablylarge, your DATA might comprise the proportion of respondents who chose A, the proportion who chose V, and the proportion who chose McGurk. That's three dependent variables, which can be viewed as interval scale for large enough N (and can be treated by logistic regression for whatever N), and for which you are interested in finding out whether the proportion in each case is a function of your cuing procedures. The only statistical test I can find to use on nominal data is a chi square, is this right? In brief, no. My textbook says that I can only use a chi square if each participant only contributes one piece of data, and only participates in one condition, which is unrealistic in a study like this, Depends on which kind of "chi square" analysis you're referring to. so can I use chi square, and if not, what should I use? Sorry; no time for further comment just now. Perhaps one or more colleagues will assist. -- DFB. -- Donald F. Burrill[EMAIL PROTECTED] 348 Hyde Hall, Plymouth State College, [EMAIL PROTECTED] MSC #29, Plymouth, NH 03264 (603) 535-2597 Department of Mathematics, Boston University[EMAIL PROTECTED] 111 Cummington Street, room 261, Boston, MA 02215 (617) 353-5288 184 Nashua Road, Bedford, NH 03110 (603) 471-7128 = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: logistic regression/error in binary variable
Konrad Halupka [EMAIL PROTECTED] wrote: Rich Ulrich wrote: I have several variables (X1, X2...) measuring various traits of individuals and one variable (Y) which is binary (survived/did not survive). I would like to check if the variation in survival can be explained with Xi variables. It looks like a typical logistic regression problem. However it bothers me that the Y variable has a non-random error. The group "survived" surely consists of individuals who *did survived*, but the group "did not survive" is likely to include some individuals which actually survived but were not detected by an observer. How to proceed with the analysis? How to proceed? -- just, proceed. Do you have any choice? Do you have some data in hand that you have not mentioned? There is no useful way to weight the data, if that is what you are wondering: a regression on a dummy-variable scored 0/1 gives you the same test as if it were scored as an (equivocating) 0/ .5. And the coefficients are easier to understand in the first one. When you describe your prediction equation, you might want to use a score other than the computer-program's default cutoff to describe the fit of prediction and outcome. But that is often the case. Do you have a hint about who covertly survived (which might suggest using a 3-group classification)? Do you have an extremely high rate of success, so that someone might be relying on the accuracy of your predictions for some purpose? Thanks for response. Indeed, I was concerned if there were some methods of weighting the data. Birds are marked and after 12 months the observer attempts to find them again in the field. Of course it is impossible to search a very wide area. Those individuals who "covertly survive" have a tendency to disperse farther than those who survive "overtly" (i.e. can be relocated within a reasonable distance from the place where they were originally found). IMO, this is an interesting problem, which is probably quite general. I encountered it in my area (hydrometeorology) some years ago: Trying to predict occurence of thunderstorms over an area based on remote sensing and large scale variables. A regression model is fitted based on the reports from a number of ground stations that are sparcely distributed over the area. The specific feature of such ground observations is that, if a thunderstorm was reported, then it occured (almost surely), however, a lot of them are simply not noticed (e.g., occured too far from any station). The applications known to me simply ignore this specific deficiency of the observations. I wonder whether it could be tractable? Of course, you need some quantitative estimate of the error. Weighted regression seems to me a correct way to account for the fact that some observations are less reliable than the others. Thus, the comment by Rich is confusing for me. Am I wrong? Are there more sophisticated ways to include such error information in a predictive model? BTW: I am new in this NG. Is s.s.e. a good place to discuss questions like the above, or the s.s.m., or s.s.c. would be better? Regards, Greg ___ Grzegorz Jan Ciach http://ia.net/~gciach = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Beta joint distribution
Hi, I have two variables with the same beta distribution (r, n-r+1). I know that the joint distribution should be a beta distribution. Is there a general equation like for the mean and the variance??? I think that the joint is a beta(r, 2(n-r+1)). I would like to know if I'm wrong... Thank you very much, Isabelle [EMAIL PROTECTED] = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =
Re: Software (fwd)
alan ... here are a few quick urls about this ... ( i don't know what it is either but, http://www.google.com ...seems to know) http://www.freebsd.org/ http://www.freebsdmall.com/ http://www.wccdrom.com/titles/freebsd/freebsd.phtml http://people.freebsd.org/~picobsd/ http://www.nzfug.nz.freebsd.org/ ... one from new zealand! At 09:06 AM 11/30/00 +1100, Alan McLean wrote: Hi Bob, What is FreeBSD? Alan Bob Hayden wrote: - Forwarded message from Ken - Of course you'll get what you pay for. "Comet" [EMAIL PROTECTED] wrote in message [EMAIL PROTECTED]">news:[EMAIL PROTECTED]... I search a good and free:) sofware of stat - End of forwarded message from Ken - When I installed Win98 on my computer at home it crashed multiple times per day. I'm writing this on a FreeBSD system that supports thousands of users and crashes less than once a year. _ | |Robert W. Hayden | | Work: Department of Mathematics / |Plymouth State College MSC#29 | |Plymouth, New Hampshire 03264 USA | * |fax (603) 535-2943 /| Home: 82 River Street (use this in the summer) | )Ashland, NH 03217 L_/(603) 968-9914 (use this year-round) Map of New[EMAIL PROTECTED] (works year-round) Hampshire http://mathpc04.plymouth.edu (works year-round) The State of New Hampshire takes no responsibility for what this map looks like if you are not using a fixed-width font such as Courier. "Opportunity is missed by most people because it is dressed in overalls and looks like work." --Thomas Edison = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ = -- Alan McLean ([EMAIL PROTECTED]) Department of Econometrics and Business Statistics Monash University, Caulfield Campus, Melbourne Tel: +61 03 9903 2102Fax: +61 03 9903 2007 = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ = == dennis roberts, penn state university educational psychology, 8148632401 http://roberts.ed.psu.edu/users/droberts/drober~1.htm = Instructions for joining and leaving this list and remarks about the problem of INAPPROPRIATE MESSAGES are available at http://jse.stat.ncsu.edu/ =