Not very good at statistics, bets I did was "Introductory Econometrics" by Ramanathan. Still got a hard question.... Suppose I want to explain number of deaths of time from a fixed pool of people (got ur attention? :) First I assume this is a binomial process. From the data I have collected over time, on average 5% dies within a period. Of coure, this number grows as the people get older. But it may also change as some start or stop smoking, using drugs, etc. So, there are explanatory variables that I can use to better predict deaths per period (and age is one of them, which changes over time). I want to specify a model that takes into account these individual variables, but optimizes the explanation of the total death rate per period. How would I do that? Normal regressions would minimize ESS with respect to predictions per sample (person per period). But basically, the "sample" I am interest in is the death rate per month. (Would that even matter aside from probably different R-squares since the TSS and ESS are different? I mean could estimator differ when trying to do what i think I am trying to do?) I guess it is a panel data analysis problem with binomial dependant variable. I am not very mathematically good, but I can dig a precise conceptual explanation with some math in it. Kind regards, Umfriend AKA Crass P.S. Please help. A referal to a book dealing with panel data would be great too. Sent via Deja.com http://www.deja.com/ Before you buy. =========================================================================== This list is open to everyone. Occasionally, people lacking respect for other members of the list send messages that are inappropriate or unrelated to the list's discussion topics. Please just delete the offensive email. For information concerning the list, please see the following web page: http://jse.stat.ncsu.edu/ ===========================================================================