New member

[Menasbo Tesfay]

Interpreting Conditional Mixed process results


Dear Statalist

I have a system of equestions with four recursive types. The first endogenous dependent variable is a binary dummy, the Second dependent variable is censored, the third is continuous and the fourth is categorical. The simple representation of the model is as follow.
1. A = f(T.C)
2. R= f(A, X’)
3. Y= f(R,X’)
4. M= f(Y, X’)

X' is vector of explanatory variables and can vary in the respective equations
Cmp ( A= T,C) (R= A, X’) (Y=R, X’) (M= Y, X’),ind( $cmp_probit $cmp_censor $cmp_cont $cpm_catago)

I got the mixed regression output ( four estimation output). My problem is how to interpret the results.For example, I want to test the effect of A on M, then how can I interpret
I would be grateful if anyone can help me.

[Phil Bromiley]

Welcome to the forum.

You did not get a quick response. You will increase your chances of a response if you follow the suggestions in the FAQ (provide Stata code, Stata output, data using dataex, etc.).

Since your cmp statement uses a bunch of macros and you have not told us what those macros are, no one can really know what you ran. The first place to look would be the cmp documentation. Since it is user written, I'm not sure whether all the normal Stata post-estimation things work with it. An alternative would be to do this in SEM or GSEM - you might find the documentation answers your question more directly.

Someone asked about a routine called KHB which might be relevant to your issues.

[Menasbo Tesfay]

Dear Statalist
Thanks for the reply. Sorry for not clarify on the variables initially. The letter A refers Access to small-scale irrigation program. The letter R refers, reliability of irrigation water on plots locate at distance from the water source ( non falling in watering days). Y refers crop productivity and M is the intensity of out market participation (Marketed surplus). X' vector of predetermined variable (Household level variables). T is village dummy where a given household resides. C is the command area (irrigation area). I hope, this makes easy to capture the relations between the variables: I will appreciate any one helping how to interpret the mixed procedure results.
Best

[Marcos Almeida]

Hello Menasbo,

Since you have recently arrived in the Stata Forum, I kindly recommend to take a look at the FAQ advices. They will help you to improve chances of getting insightful replies to your queries

For example, the (appropriate) title is really important, because it will call the attention to a given topic, and fellows who have expertise on it will probably feel enticed to join in.

What is more, with a uninformative title, shall the thread get lots of helpful comments and useful commands, it won't be easily found later by people with similar difficulties.

Therefore, in the forthcoming "new topics", instead of the quite uniformative "New Member", just for the sake of giving an example, you may choose something like "Interpreting Conditional Mixed process results".

Hopefully that helps.

[Menasbo Tesfay]

Thanks Marcos for kindly adivise.
Best regards,
Menasbo