I am doing some analyses using the NHANES III data set. NHANES is a national survey using a nested stratified sampling design for which the routines available in SUDAAN are recommended for analysis.
I need to update and correct some of the conclusions that I posted about a week ago about the cox.zph function, at least about the version of cox.zph in SPlus for Windows 6.2. (I have uninstalled 6.
Another question for the coxph gurus: At the end of the print out of summary(coxph) there is a line that reads: "Rsquare= 0.316 ( max possible= 0.97 )" (Numbers vary, of course, depending on the mode
Terry and all: I appreciate deeply Terry's point about it not being clear what reduction of unexplained variability means in a Cox analysis, and that there are not unique definitions of this concept.
HI, folks. I'm back again. I am trying to get predicted values for a Cox model in which the two independent predictors are both factors with four levels. I am after not only the separate levels, but
Thanks, Terry Larry Hunsicker Larry H asks: Why do none of the predicted values from a coxph model, using factors, =0? The "reference" group for coxph is a fictional subject with the MEAN value for e
I'm back to the trough, folks. I want to make plots for which I don't have a name but which I shall describe. One wants either to show the estimated odds ratio for a treatment, or a risk factor, or w
I have a Cox regression problem that I am working on for which I have replicate measurements of one of the predictors, a predictor which is clearly measured with error. Are there any routines in (or
I'm back to the trough with a very specific problem. I should like to examine the dfbeta residuals of a multiple-event cox model that includes a cluster term. One of the covariates in the model is "
Thanks to Terry Therneau for a rapid and effective solution to the problem of computing a dfbeta residual. He notes that the problem probably results from an error when the residual function tries to
I have found an anomaly when one uses the parameters control and correlation simultaneously in the function lme. I have illustrated this in the appended text file, using the Orthodont example. You w
example given in the book fits a random intercept for both Sex and age. That is done by making random=age|Subject You ask about fitting an AR(1) structure for the residual errors within Subject. To
I have been trying to do an lme analysis of a moderately large data set (1,715 patients with 11 measurement times). For various reasons I need to stipulate an unstructured correlation matrix, which