tag:blogger.com,1999:blog-8269737761286917895.post946516522752215110..comments2024-02-19T23:07:54.532-08:00Comments on Coffee and Econometrics in the Morning: Porting Stata-like Marginal Effects to LaTeXTony Cooksonhttp://www.blogger.com/profile/12565713889808330198noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-8269737761286917895.post-76329660438296373142012-04-03T06:19:36.381-07:002012-04-03T06:19:36.381-07:00Hi, I'm a French PhD student an I work on a to...Hi, I'm a French PhD student an I work on a tobit model with the 2steps Heckman method (function "heckit2fit" from the "sampleSelection" package). I need to calculate the marginal effects for the selection equation (probit) and for the outcome equation. For the selection equation I found a mfx function on a working paper that seems to work well (http://www.ucd.ie/t4cms/WP11_22.pdf) but for the outcome equation I find nothing that could give the marginal effects of variables that are used in both equations as Stata does. Can you help me with that? <br />Another point : Whether I use the mfx fucntion decribed in the working paper I mentionned above or my own script for the calculation of the probit marginal effects, I find similar marginal effects... but they both are 10 times inferior to the marginal effects given by stata (I had the opportunity to compare thanks to a friend who has Stata)...I don't understand where these differences come from, the only thing I notice is that "rho" of my heckit model is absolutely not the same with Stata and with R... <br />Thanks you for your help. <br />Laurence Delattre. <br />(French National Institute for Agricultural Research)Anonymousnoreply@blogger.com