information on Stata for people on old versions [was: Re: st: Questions ... SEM]

[Nick Cox <njcoxstata@gmail.com>]

What Yuval missed for some reason no doubt could be missed by others.

Let's extend it:

1. If you are using some version of Stata before the present one
(12.1), and wish to know what's in the current release, then visit
http://www.stata.com/stata12/ for a good start.

2. The bottom of every Statalist posting exemplifies the fact that
much of the help for Stata 12 is also available over the internet. The
example is given of

http://www.stata.com/help.cgi?search

The recipe is easy: once you know the name of a manual entry, you can
replace "search" with that name. I just guessed at

http://www.stata.com/help.cgi?sem

and it does work. (No; neither the .pdf manuals nor the software is
available for download.)

3. If you want to know what's been written in the Stata Journal,
typing -search- with keywords in your Stata will tell you to the
extent that your Stata is up-to-date. But a search from

http://www.stata-journal.com/archives/

will work for anyone. Papers published more than 3 years ago are
accessible to all in .pdf form.

4. If you want to know what's in the next release. or indeed when that
will be, the historical pattern is that Bill Gould will announce this
on Statalist when both content and date are certain (and not before) .
For last time, see

http://www.stata.com/statalist/archive/2011-06/msg01146.html

I say "historical pattern", as I have no idea whether StataCorp will
do something new next time.

Nick

On Thu, Nov 1, 2012 at 6:27 AM, Yuval Arbel <yuval.arbel@gmail.com> wrote:

> Regarding my previous question I found the following link in stata website:
>
> http://www.stata.com/stata12/structural-equation-modeling/explanation/
>
> This link answers my previous question. However, just to be sure:
>
> Does the example given there imply that the variable X is unobserved,
> namely does not included in the dataset?

On Thu, Nov 1, 2012 at 6:55 AM, Yuval Arbel <yuval.arbel@gmail.com> wrote:

>> William, unfortunately I don't have stata 12. However, my question is
>> conceptual. Previous stata versions had the possibility to estimate
>> system of equations using, for example, -reg3-. The question is what
>> is the supplements in
>> -sem-. The broader question will therefore be whether the terms
>> "Structural Equation Model" and "System of Structural Equations" are
>> identical

On Thu, Nov 1, 2012 at 6:38 AM, William Buchanan

>>> Prior to Stata 12 there wasn't any native SEM functionality. If you look at [SEM] on the online help or the documentation itself you can find additional details regarding what you might be looking for.

On Oct 31, 2012, at 21:29, Yuval Arbel <yuval.arbel@gmail.com> wrote:

>>>> William, thanks for the quick answer. Is there any related material in
>>>> stata journal? what are the supplements (if any) of the -sem- commands
>>>> with respect to previous versions?

On Thu, Nov 1, 2012 at 6:21 AM, William Buchanan
>
>>>>> The -sem- command was introduced in Stata 12. You can find user written packages that can fit measurement models, path analysis models, and -gllamm- is incredibly flexible.

On Oct 31, 2012, at 21:08, Yuval Arbel <yuval.arbel@gmail.com> wrote:

>>>>>> Can anyone address me to a relevant stata material regarding
>>>>>> structural equation models?
>>>>>>
>>>>>> From my previous knowledge,, structural equations are the original
>>>>>> equations in a system. Is there more to it?
>>>>>>
>>>>>> I'm working with stata 11.2. Interestingly, I cannot find or install
>>>>>> the -sem- command. On the other hand, I can install other related
>>>>>> commands,
>>>>>> such as, -lmasem-. Is there any particular reason for this?
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/faqs/resources/statalist-faq/
*   http://www.ats.ucla.edu/stat/stata/