SEM: estimating curvilinear effects

[Sebastian Koehler]

Hello,

I need some advice on whether what I want to do is nonsense because I didn't find previous examples on this. I'm running an SEM with a linear and quadratic term (X and X2) and a continuous mediator (M) on Y. With -estat teffects- I can get the indirect effects of X and X2. However, I'd like to learn about their joint indirect effect. What I see in the output is that there is a direct effect of X on Y but not an indirect effect via M, while there is no direct effect of X2 on Y but an indirect one via M.

I tried -sem, coeflegend- followed by -nlcom- like in:
nlcom ([M: b_X] * [Y: b_M]) + ([M: b_X2] * [Y: b_M])

Best wishes,
Seb

[Phil Bromiley]

You didn't get a quick answer. You'll increase your chances of a useful answer by following the FAQ on asking questions - provide Stata code in code delimiters, readable Stata output, and sample data using dataex.

We don't know exactly what model you ran, and what the problem with the nlcom is.