binary_mediation command

[Marie Gabel]

Hello,

Using Stata 16, I am trying to run an existing do-file again in which I used the binary_mediation command.

Unfortunately, the command is no longer working. The following error occurs:


binary_mediation, dv(REG) mv(IPC) iv(sz3) cv(rATT sz2 sz1)
command binary_mediation is unrecognized
r(199);

Also using help or search functions do not lead to any results.

Is the binary_mediation command still existing?
Or, if not, is there an alternative way of calculating binary mediation using Stata 16?


Many thanks in advance

Best regards

Marie

[William Lisowski]

This recent topic suggests that the command is no longer made available by UCLA's IDRE, which this community contributed command originated.

https://www.statalist.org/forums/for...nary-mediation

Perhaps someone else can explain if you can get what you need using Stata's sem or gsem commands.

[ericmelse]

Here is the ado and sthlp file, version version 2.4 -- 10/14/11.
I downloaded them from the UCLA's IDRE server some time ago.
I have also attached a PDF of the webpage that discusses the command binary_mediation.
As mentioned in the post, mentioned by William (#2), note that this page is no longer available at UCLA's IDRE server but for more on mediation, from them, you can go here (section mediation).

But, maybe, you might also find this paper relevant to this subject:
The Stata Journal (2005) 5, Number 1, pp. 64–82, Michael N. Mitchell & Xiao Chen. Visualizing main effects and interactions for binary logit models, and its related ado files that you can find here:

net get vibl

Best,
Eric

[Marie Gabel]

Many thanks for your quick and helpful response. Your answer and the ado file helped a lot.

Best Marie

[Josep Roman Juan]

Hi

Using Stata 14.2, I'm trying to run a binary mediation using this same command (bivariate_mediation).

However, after introducing the syntax:

binary_mediation, dv(bck) iv(i.survy) mv(i.bmi_cat) cov(sex agecat FAS)

I've been returned the following error:

/ is not a valid command name
r(199);

Any help is much appreciated; I've been through most of the binary_mediation posts on here and I didn't see anything similar, so if I missed something please point me in that direction. Thank you!

Josep

[Joanna Davies]

Hello,

Just picking up on this thread - does anyone know why the binary_mediation command is no longer available? Should we avoid using it if its no longer supported?

Am i correct in thinking that there are no currently available Stata commands that support estimation of indirect and direct effects for non-linear models (why is this the case?)?

thank you,
Joanna

[Maarten Buis]

You could look at khb

[Joanna Davies]

Thank you Maarten - looks like khb could be just what i need.