Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.
From | "Chris Ambrey" <chris.ambrey@gmail.com> |
To | <statalist@hsphsun2.harvard.edu> |
Subject | st: Average marginal effects in ordered probit models |
Date | Thu, 13 Sep 2012 17:31:49 +1000 |
Hi there, I've been struggling with this question for a while and I have discovered there are many ways to achieve the end result or something similar or close to it using margins or mfx2. However, I want to use predictnl and/or nlcom to obtain the average marginal effects for several interaction terms in the one model. I'm estimating an ordered probit model. I'm using version 11.2 of STATA. Life satisfaction is my dependent variable (0-10) and I have greenspace (continuous, percentage of greenspace in the census tract) and a number of socio-economic and demographic characteristics as explanatory variables (some dichotomous and others continuous). I've come across this http://www.stata.com/support/faqs/stat/mfx_interact.html support page which was helpful, but I still don't quite follow it. Also, I'm familiar with the inteff command. I haven't been able to get it to work on my computer (Windows 7), but I understand it really only handles one interaction term. I'm also aware of the various user written commands provided by Long and Freese (2006). Although, I don't think these commands actually provide the average marginal effects but the marginal effects at different values for the explanatory variables. Please correct me here if I'm wrong here. Furthermore, I have found a very recent article "Interaction Terms in Nonlinear Models" by Karaca-Mandic, Norton and Dowd (2012) which has been very helpful. Still, I'm not sure about the syntax for an ordered probit model in STATA using predictnl and/or nlcom to obtain the average marginal effects. I would like to have the standard errors (obtained using the Delta Method) and the p-value etc... as well. Additionally, do the average marginal effects apply to a particular outcome, such as a self-reported life satisfaction score of 10 and are thus interpreted as the probability of reporting say a 10 for a one unit increase in the explanatory variable? I would certainly appreciate any assistance or direction in this regard. I've been wrestling with this issue for some time. Kindest regards, Christopher Ambrey * * For searches and help try: * http://www.stata.com/help.cgi?search * http://www.stata.com/support/statalist/faq * http://www.ats.ucla.edu/stat/stata/