[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

RE: st: Why no predicted probs but exponentiated linear prediction for -adjust- after -xtgee- or -xtmelogit-?

From   "Verkuilen, Jay" <>
To   <>
Subject   RE: st: Why no predicted probs but exponentiated linear prediction for -adjust- after -xtgee- or -xtmelogit-?
Date   Thu, 18 Dec 2008 17:17:28 -0500

Nick Cox WROTE:

>The absence of a dialog means no more than that. Neither SSC nor SJ
insist on dialogs in submissions: if either did, the number of
submissions would fall drastically. <

No doubt about it. I've written a GUI. Even with modern software
development tools it's not easy.

>In Stataland, "official" has a bespoke sense of whatever software or
documentation or opinions or mugs or T-shirts emanate from StataCorp. <

Precisely. I'm getting in the mind of a fearful user. These people are
paranoid of "user contributions" and they really like point and click.
Software like R is attractive because of the price, but the blizzard of
user contributions without a corporation standing behind them puts many
off (as does the general difficulty of getting data into R). 

Stata is actually a really good choice for them because it's got lots of
what they need and more and is set up in the way they understand; most
crucially it has a great GUI. They'll come around eventually but
"official" and dialog boxes are very helpful sales entry points.
-adjust-, being a very powerful way to generate output from many models,
strikes me as a potential "killer app." *We* all know it's a wrapper for
-pred- but the guts of Stata are too much for the start. 

I like Stata and will buy it no matter what happens, and therefore my
students will use it in class. I'm just throwing out some feedback on
what I think would make Stata really attractive to users, many of whom
are POed right now by SPSS' arbitrary and expensive licensing. 


*   For searches and help try:

© Copyright 1996–2017 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index