Moderated mediation analysis with 2 mediators

[Ava Buxton]

Hi everyone,

I am conducting a moderated mediation analysis with 2 mediators

How can I create interaction terms for the following model?:

Thank you.

[Jeremy Miles]

Is W continuous or categorical?

--
You received this message because you are subscribed to the Google Groups "lavaan" group.
To unsubscribe from this group and stop receiving emails from it, send an email to lavaan+un...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/lavaan/CAFUb6P9f_W0i2Rp2mXTsHQywwGb0CxxXvciyTnrS4S6w1LYZhg%40mail.gmail.com.

[Ava Buxton]

Continous

To view this discussion visit https://groups.google.com/d/msgid/lavaan/CAMtGSxnuporUcYFHk%3DYkag6i7eAFZ8Sqj7aoNdreAO0q8v6n0A%40mail.gmail.com.

[ahmad valikhani]

Interaction between X and W = X scores * W scores 

Interaction between M and W = M scores * W scores

To view this discussion visit https://groups.google.com/d/msgid/lavaan/CAFUb6P9aAMKkMHVyP0u7JxqGgYgy8LA2Lx-iXEesy2Y4T5xH3Q%40mail.gmail.com.

[Andries Masenge]

mean centered X and W scores.

---

From: lav...@googlegroups.com <lav...@googlegroups.com> on behalf of ahmad valikhani <valik...@gmail.com>
Sent: Wednesday, July 01, 2026 12:44
To: lav...@googlegroups.com <lav...@googlegroups.com>
Subject: Re: Moderated mediation analysis with 2 mediators

To view this discussion visit https://groups.google.com/d/msgid/lavaan/CAOZ3ZRgYd%2B24W%2B7TQLfj6oOeenKC9u6-OExwF6%3DX96VrAQL1Yw%40mail.gmail.com.

This message and attachments are subject to a disclaimer.
Please refer to http://upnet.up.ac.za/services/it/documentation/docs/004167.pdf for full details.

[Ava Buxton]

Thank you!

Would I mean center all variables except y?

Since I have 2 mediators, do the following interaction terms make sense?:

 Interaction between X and W = X scores * W scores

 Interaction between M1 and W = M1 scores * W scores
 Interaction between M2 and W = M2 scores * W scores

Is there no term needed that includes y?

To view this discussion visit https://groups.google.com/d/msgid/lavaan/GV1P193MB22626EBFBA5FBD2AB2112EABACF62%40GV1P193MB2262.EURP193.PROD.OUTLOOK.COM.

[Christian Arnold]

Hi Ava, 

please think twice before relying on scores for testing moderation in SEM. Whatever is being recommended here, both the product‑indicator approach and the newer SAM framework offer measurement‑consistent ways to estimate latent interactions in lavaan. A recent BRM paper (https://link.springer.com/article/10.3758/s13428-024-02532-y) shows how to apply SAM for moderarion. If you’re testing moderation with latent variables, PI or SAM will give you a much more defensible solution.

Best, Christian 

---

Von: lav...@googlegroups.com <lav...@googlegroups.com> im Auftrag von Ava Buxton <avatb...@gmail.com>
Gesendet: Donnerstag, Juli 2, 2026 8:38:29 PM
An: lav...@googlegroups.com <lav...@googlegroups.com>
Betreff: Re: Moderated mediation analysis with 2 mediators

To view this discussion visit https://groups.google.com/d/msgid/lavaan/CAFUb6P_%2BknNfQrZzNUvB2juZNjpF0aqGpKxrW27S1URvNon8tQ%40mail.gmail.com.

[Ava Buxton]

Thank you. Does anyone have advice for bootstrapping? I see variations of code for this but am confused about how to approach it in my case. My model looks like Hayes Model 15 with 2 mediators (all continuous).

To view this discussion visit https://groups.google.com/d/msgid/lavaan/AM6PR04MB595761904F52650FF938DBE5FCF52%40AM6PR04MB5957.eurprd04.prod.outlook.com.