I was gone this weekend, and Stata 11 was waiting for me upon my return.
Early impression: in defiance of the law of diminishing returns to higher version numbers, this is a major upgrade, perhaps comparable to if Stata had skipped version 9 and gone straight from 8 to 10.
1. For people interested in what their regression results actually mean, the margins command is a sweeping and elegant improvement.
2. The multiple imputations module makes their use much easier. While multiple imputations are not All That, social science would benefit from an increased mundaneity of their use.
3. The do-file editor looks great so far. If you are doing any kind of complicated data analysis and still aren’t using something with syntax highlighting, well, YOU ARE MAKING YOUR LIFE MORE DIFFICULT AND ERROR PRONE THAN IT NEEDS TO BE.
3 thoughts on “stata 11: the early review”
“For people interested in what their regression results actually mean, the margins command is a sweeping and elegant improvement.”
why is this necessary? i thought Stata has always reported p-values
Looks promising. How does the new do-file editor compare with using an external editor? I have been thinking about switching from Textpad/Notepad++ to vim.
Among the many new features advertised in Stata 11, one very convenient bonus is the addition of pdfs of all the manuals at no added cost.