Re: More fun with E(X*Z^2)
In article <8a1r90$qgs$[EMAIL PROTECTED]>, Enfilade <[EMAIL PROTECTED]> wrote: >Hi, > Many thanks to Herman Rubin for his comment on my earlier post. I >may not have fully understood him, so I'm restating the problem with a >little more detail. I also progressed a bit further, and am wondering >if it is, in fact, as far as I can go. > The problem is to derive an expression for E(X*Z^2). If you are >wondering why, it is part of the formula for E(dx*dz^2), where dx=X - E >(X), and dz = Z - E(Z). I know that if BOTH X and Z are continuous and >distributed bivariate normal, E(dx*dz^2) = 0. I'm trying to determine >what this odd moment will be if X is binary 0/1 with proportion (px) >free to vary, and Z distributed univariate normal. (Note: I will be >very pleased if someone tells me E(dx*dz^2)=0 here, too. I believe it >is the case with px=.5, but only that case.) Not even here. To show the problem, let me assume that Z is normal with mean 0 and variance 1, and P(X=1) = .5 and X is uncorrelated with Z. This is not enough to determine E(X*Z^2). To see this, note that if X and Z are independent, the expectation is .5. For another model, assume that P(X=1|Z) = exp(-1.5*Z^2). This makes Z normal with mean 0 and variance .25 conditional on X=1, and also P(X=1) = .5. However, E(X*Z^2) = .125. It is the multidimensional normal which is the unusual case. -- This address is for information only. I do not claim that these views are those of the Statistics Department or of Purdue University. Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399 [EMAIL PROTECTED] Phone: (765)494-6054 FAX: (765)494-0558 === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Cox & Wermurth short course
OTTAWA SHORT COURSE ON APPLYING STATISTICS 2000 Monday & Tuesday, May 15&16 2000 9am to 5pm Ottawa, Ontario Canada Sponsored by Department of Epidemiology and Community Health of the University of Ottawa and the Loeb Health Research Institute at The Ottawa Hospital MULTIVARIATE DEPENDENCIES Lecturers: David Cox, Nuffield College, Oxford and Nanny Wermuth, ZUMA - Center for Survey Research, Mannheim Registration Fee: $300.00 (Cdn) - The fee to Graduate Students is $150.00. Includes the lecturer's book "Multivariate Dependencies - Models, Analysis and Interpretation", Chapman and Hall, 1996 The aim of the course: The course will provide a systematic discussion of a basis for the analysis and interpretation of complex multivariate data. The course is designed for those making extensive use of statistical methods in health and social science research who need to meaningfully interact with both other statisticians and scientists / decision makers. Common issues and themes that arise in almost any form of careful data analysis, complex or otherwise, will be discussed. The course is intended for those concerned with the analysis and interpretation of complex data, especially but not entirely observational data. The applications from the social and health sciences are wide-ranging including, for example, studies of medical interventions and of sociological or psychological development. While the primary emphasis will be on statistical methods for direct use in applications, some issues of theoretical interest will also be addressed. One central theme will be the role of independence graphs and on processes by which the data could have been generated. This theme suggests the recasting and restructuring of fairly standard and well known statistical techniques to better focus and report on important data analysis issues. No software presentations will be given, the emphasis being largely on methods which can be implemented within standard packages. A fairly similar course was given last Fall at the Fields Institute in Toronto (http://www.fields.utoronto.ca/cis-scourse.html#MARKOV). While the course is based in part on the lecturers' book "Multivariate Dependencies - Models, Analysis and Interpretation" (Chapman and Hall, 1996), a number of important developments since the book will be described. As well as core material a number of specific research questions will be discussed in detail; the corresponding data can be obtained via the internet (soon at http://www.lri.ca/). For further information and registration email the course organizer, Keith O'Rourke at [EMAIL PROTECTED] === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Re: More fun with E(X*Z^2)
In article <8a2sv8$[EMAIL PROTECTED]>, [EMAIL PROTECTED] (Herman Rubin) wrote: > > Not even here. To show the problem, let me assume that Z > is normal with mean 0 and variance 1, and P(X=1) = .5 and X > is uncorrelated with Z. This is not enough to determine > E(X*Z^2). > > To see this, note that if X and Z are independent, the > expectation is .5. For another model, assume that > P(X=1|Z) = exp(-1.5*Z^2). This makes Z normal with > mean 0 and variance .25 conditional on X=1, and also > P(X=1) = .5. However, E(X*Z^2) = .125. > > It is the multidimensional normal which is the unusual case. > -- Thanks. I think I undertand it now. To determine (X*Z^2), it appears I either have to specify the variance of Z in both categories of X (getting overall V(Z) from the variance formula for mixture models), or assume they are equal. Leaving in two variance terms doesn't complicate the expression too much, so I'll probably just let the user make the homogeneity assumption if they wish. -E Sent via Deja.com http://www.deja.com/ Before you buy. === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Descriptions of Rates
QUESTION: Is there a good reference on reading rates??? Milo In describing rates as part-whole ratios, it seems that the part is often introduced by 1. "rate of" [ rate of deaths, rate of births, etc.] or 2. a modifier of "rate" [death rate, birth rate,etc.] Using a relative clause seems feasible. 3a. rate of {part} who/which are due to {part}. [death rate due to guns]. 3b. {part} rate who/which are {part} [death rate due to guns] Now consider combinations of 1 & 2. In some cases, the adjective does not indicate the part. 4a. a fixed/higher/growing rate of interest 4b.a nice/tiny/adequate rate of interest In some case, the adjective indicates the part while the rate of" phrases introduces a whole or a condition. 5a.death rate of males 5b.birth rate of young females In some cases, the adjective indicates the whole or condition, while the "rate of" introduces the part. 6a.males' rate of accidents 6b.teenage rate of pregnancies In some cases, the adjective simply conditions: 7a.a commercial rate of interest In some cases, we can get three different influences: 8a.the highest unemployment rate of the nation's 11 industrial states 8b.the high divorce rate of their parents generation === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Newbie question - "apply" hangs SPLUS
Hi there, I'm new to SPLUS so this question my sound silly to you. I have a big matrix (dim about 25,000 by 20) which was of mode numeric, but by some accident (probably me trying out some things) has converted to mode list, that is, both rows _and columns_ are lists of numeric data. Now that's not what I need to have, so i'm trying to get it back to the original format using something like new.matrix.name<-apply(matrix.name,2,unlist) It works fine if I try this for the, say, first 10 lines only - new.matrix.name<-apply(matrix.name[1:10,],2,unlist) - but SPLUS seems to hang when I go for the big thing. It works (or looks like it's working) for hours and hours (more than 14 hrs once) and then I have to kill it. I'm using SPLUS 2000 an a WinNT Workstation 4.0 machine with a Celeron 300 MHz CPU and 64 MB RAM. Pagefile max size is set to 225 MB. Please help me out with this since getting this operation done is critical for the work i have to do. TIA, Andrei === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
job announcement
Michigan State University College of Agriculture and Natural Resources POSITION ANNOUNCEMENT POSITION: 12-month Tenure-track Assistant Professor LOCATION: Michigan State University College of Agriculture and Natural Resources (CANR), Department (plant sciences) to be determined. CLOSING DATE: May 15, 2000 or until position is filled. RESPONSIBILITIES: (1) Teaching (50%) The successful candidate is expected to teach two senior/graduate level courses in statistical methods in the plant and soil sciences and an advanced graduate course in spatial statistics (alternate years) or a related field. (2) Research (25%) Develop independent and collaborative research in biometry directed at the plant and soil sciences. Develop statistical applications to solve problems in agriculture and natural resources. Publish results in appropriate scholarly journals. Pursue extramural funding for research. Direct graduate student research. (3) Collaborative Research and Consulting (25%) Serve as statistical expert on research teams within the CANR. Provide statistical consultation to faculty and graduate students in the Department and the CANR. Serve on graduate committees. Serve as co-director of the CANR Statistical Consulting Center. [www.fw.msu.edu/orgs/canr_big] QUALIFICATIONS: Ph.D. degree in a plant science discipline and/or Applied Statistics with a specialization in spatial statistics or a related field. Experience with generalized and nonlinear mixed effects models is desired; expertise with modern and intensive computing methods is required. If the Ph.D. is in Applied Statistics, strong emphasis in a plant science discipline is required. The individual must be an effective communicator with the ability to teach statistical methods to graduate students in the CANR. The individual must also have a strong interest and commitment to statistical consulting. The successful candidate must further demonstrate ability to conduct research in spatial statistics as applied to inferential problems in plant and soil science-related fields. APPLICATION: Candidates should send a cover letter detailing their interest and qualifications for the position, curriculum vitae, academic transcripts, a brief statement of teaching and research interests (not to exceed two pages), and three letters of recommendation to: Dr. Carl W. Ramm, Search Committee Chair[EMAIL PROTECTED] Department of Forestry Michigan State University East Lansing, MI 48824-1222 MSU is an Affirmative Action/Equal Opportunity Institution
Advance registration-Stat&Health Conference, June 2000
- STATISTICS AND HEALTH: Edmonton Statistics Conference June 11-13, 2000, Edmonton, Alberta Call for Advance registration by May 1, 2000 - Organized by the Biostatistics Research Group, Statistics Centre, Department of Mathematical Sciences at University of Alberta, Edmonton, the conference aims to provide a catalyst toward understanding the role of statistical sciences in scientific research, facilitating communication among relevant disciplines. The featured speaker is Bradley Efron (Stanford) on "BOOTSTRAP BIOSTAT". The invited speakers are as follows: * CORRELATED DATA IN HEALTH RESEARCH: Rao, JNK (Carleton); Heagerty, P (Washington); Wang, MC (Johns Hopkins) * META-ANALYSIS AND ERRORS IN ESTIMATION: Berlin, J (Pennsylvania); Greenland, S (UCLA); Moher, D (Ottawa) * SPATIAL MAPPING OF DISEASE RISKS AND SMALL AREA ESTIMATION: Dean, C. (Simon Fraser); Waller, L (Emory); Walter, S (McMaster) * HIERARCHICAL MODELS-NEW DEVELOPMENTS: Lawson, A (Aberdeen); McCulloch, C (Cornell); Normand, S-L (Harvard) * COST, COST EFFECTIVENESS AND DECISION ANALYSIS: Bryant, H (AB Cancer Board); Manning, W (Chicago); Philipson, T (Chicago) * EPIDEMIOLOGY AND METHODS: Gentleman, J (Nat'l Center For Health Stat); Liang, KY (Johns Hopkins); Morgan, K (McGill)/Bull, S (Toronto) For more information, visit the conference website at www.stat.ualberta.ca/~brg/conf.html or inquire at [EMAIL PROTECTED] To qualify for the early discount rates, registrations must be postmarked, faxed or emailed before May 1, 2000. === K.C. Carriere Associate Professor of Statistics Department of Mathematical Sciences University of Alberta (tel)780-492-4230 Edmonton, AB T6G 2G1(fax)780-492-6826 Home: www.math.ualberta.ca/~kcarrier/kcarrier.html Visit:www.stat.ualberta.ca/~brg === === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
How do you compute the probability a paker hand that is void in at last one suit.
I have an assignment where I have completed 5 out of 6 question relating to probability and counting. I'm stuck on the following problem: Five cards are selected at random without replacement from a 24 card deck (ie. nines through aces). What is the probability of a void in at least one suit. Example:What is the probability of a full house: Answer: [c(6,1)*c(4,3)*c(c,1)*c(4,2)]/c(24,5) Any help or hint that you may have would be greatly appreciated. Sincerely, Aaron Moynahan === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===
Re: A question on hypothesis testing wrt a factor in a GLM model.
Thank you for your reply, the only one I have got so far. "Donald F. Burrill" wrote: > > Looks to me like a simple typo. The null hypothesis is > H0: beta1 = beta2 = 0 > and I would attribute the "+" sign to a typing error ("+" is a shifted > "=" on most keyboards). You may be right here. This problem is found in a survival analysis text. The text does have a few obvious typos. Hopefully this is just another instance. > While the _interpretation_ of non-zero estimates of the betas > (should H0 be rejected) depends on the specific choice of > parameterization for the three levels, all of the choices that spring to > mind would have this null. It corresponds to the standard null > hypothesis in an ANOVA with a 3-level factor: > H0: mu_A = mu_B = mu_C, with 2 degrees of freedom. Agree. Now I am interested in my second question. That is, if I do want to test that a null hypothesis H0: beta1=0 or beta2=0, which, in my understanding, is different from H0: beta1=0 and beta2=0, I suppose the test for the former should be different from the test for the latter. Is this correct? > On Sun, 5 Mar 2000, Paul Y. Peng wrote: > > > Suppose that I have a factor with three levels A, B, and C. If it > > is used in a GLM model as a covariate, I will have two parameter > > beta1 and beta2 (assuming they are for level B and C). To test a > > statement "Any of the last two levels (either level B or level C) > > has a different effect on the response variable than level A," a text > > book says that the null hypothesis should be H0: beta1 + beta2 = 0. > > However, I cannot figure out why this is the null hypothesis. I > > thought the hypotheses should be H0: beta1=0 or beta2=0 vs > > Ha: beta1!=0 and beta2!=0. If they are correct, how should I test > > them using say Wald's test or the likelihood ratio test? Thanks > > for any hints. > > Why not the standard F test, comparing the residual SS for a model that > includes the factor to the residual SS for a model that does not include > the factor? I should say the original question is from a survival model rather than a GLM (I mentioned GLM because I believe a GLM may be more popular than a survival model). In survival models, we usually use Wald's test or LRT. Thanks. Paul. === This list is open to everyone. Occasionally, less thoughtful people send inappropriate messages. Please DO NOT COMPLAIN TO THE POSTMASTER about these messages because the postmaster has no way of controlling them, and excessive complaints will result in termination of the list. For information about this list, including information about the problem of inappropriate messages and information about how to unsubscribe, please see the web page at http://jse.stat.ncsu.edu/ ===