I'm producing a Kaplan Meier Survivor function using proc lifetest, and noticed in the example that the results are different than the book Kalbfleisch and Prentice (v2 pg16). Basically, the sample in the book has 20 observations at at the first even the function is 1/19 (remaining amount) while in SAS it appears as they calculate it as 1/20. this is based on the data in the proc lifetest sample data/documentation.
Does anyone know why this is? and What differences there are between the two methods? I don't know if a factor of 1 would be a huge deal, except if there was a sudden cutoff point where a lot of people died at once.