[Joaquin Carrascosa]

Thanks, it´s working now!

[daniel klein]

For those with earlier versions of Stata, see my workaround here.

Best
Daniel

[Richard Williams]

Like I said in that same thread Daniel links to, it would be nice if Mata had some sort of saveold function. Barring that, a FAQ on this would be good.

I have every version of Stata from 7 on. I don't see any need to delete old versions, and (unlike many programs) it is easy to copy old versions and keep them running. It is quite handy for making sure I don't unnecessarily zap my user-written programs when I make changes to them.

If authors can't compile their Mata code under an earlier version of Stata, providing the .mata files is the next best thing.

[Richard Williams]

Incidentally, when I am having trouble with a program or think it is buggy, the first thing I do is run the commands

update all
adoupdate

If I am lucky the problem has already been fixed.

Also, if a student using Stata 13 or 14 comes to me with a problem, I try to run it, or have them run it, under Stata 15. I don't want to spend hours on a problem that Stata fixed years ago. Of course, my students can always go to labs that have Stata 15, but not everyone has that option.

People occasionally write me about bugs with my programs. Sometimes I find there was a bug, but I fixed it back in 2007. If you are using a program that was in Stata Journal, you can check to see if there is a newer version on SSC.

[Joaquin Carrascosa]

For those with earlier versions of Stata, see my workaround here.

Best
Daniel

The workaruond worked perfectly on Stata 13! Thanks! I´m already working on my models (class destinations dependant on class origins, mediated by education and social capital). Up to this point i had been using SEM for path analysis models using continuos variables (occupational status and years of education)
Very helpful as always.

[Jon Hoefer]

Coming from someone with "half-knowledge", I have the following question: Mr. Kohler said in #8:

Whether or not it is correct to call the SE of the difference that -khb- shows "bootstraped" depends on whether or not the technique to derive the SE from the bootstraped VCE invokes additional assumptions. Addmitedly, I must leave this debate to the statisticians. The SE is derived using Sobel's Delta Method; see our paper on khb in SJ11-3

. Referring to that, I want to use the -khb- for a mediation analysis and then wrap a bootstrapping comment around it, as done by Vivian. I ,however, want to do it, in order to benefit from the Bootstrapping feature that allows me to make no assumption about the underlying distribution and because the -khb- command, as I understand it, uses the Sobel test, which is exactly which I do not want to use, due to reasons elaborated by Hayes (2017 https://primo-49man.hosted.exlibrisg...92471800002561). Is the -khb- tool in addition with the bootstrapping procedure hence feasible for my use case?