<div dir="ltr">Dear Peter,<div><br></div><div>It depends on what you would want to estimate, but it's certainly possible to do this under certain assumptions. For example, you can estimate the indirect effect of treatment on outcome through either mediator 1 or 2. It's also possible to estimate the indirect effect of treatment on mediation 2 through mediator 1.</div><div><br></div><div>Kosuke</div><div><br></div><div>---------------------------------------------------------<br>Kosuke Imai Office: Corwin Hall 036<br>Professor Phone: 609-258-6601<br>Department of Politics Fax: 609-258-1110<br>Princeton University Email: kimai@Princeton.Edu<br>Princeton, NJ 08544-1012 <a href="http://imai.princeton.edu">http://imai.princeton.edu</a><br>---------------------------------------------------------<br></div><div class="gmail_extra">
<br><div class="gmail_quote">On Sun, Nov 2, 2014 at 4:21 PM, Peter Ronald Belmi <span dir="ltr"><<a href="mailto:peterbelmi@gmail.com" target="_blank">peterbelmi@gmail.com</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex"><div dir="ltr">Hi Dr. Imai,<div><br></div><div>Can I use the mediation package if I have a model that's something like this:</div><div><br></div><div>Treatment -> Mediator 1 -> Mediator 2 -> DV<br></div><div><br></div><div>Is there a way to test the indirect effect in this model?</div><div><br></div><div>Cheers,</div><div>Peter</div></div>
</blockquote></div><br></div></div>