1 item has been added to your cart.
| Title | Cox regression results in Stata versus SAS | |
| Authors |
William Gould, StataCorp James Hardin, StataCorp |
|
| Date | November 1996 |
If you would like to match results from SAS PHREG, add a small number to the time variable for the deaths (so that Stata will censor the values and remove them from the risk pools for the deaths).
Neither. The fault lies with whomever collected the data. Time is continuous and the data was not collected at a fine enough resolution. When your dataset does not differentiate the order that events occur, you leave it to the software to decide for you. This is not a good idea in general.
