tag:blogger.com,1999:blog-2198942534740642384.post3958574395205072657..comments2023-10-24T03:16:41.009-07:00Comments on Econometrics Beat: Dave Giles' Blog: When Can Regression Coefficients Change Sign?Dave Gileshttp://www.blogger.com/profile/05389606956062019445noreply@blogger.comBlogger6125tag:blogger.com,1999:blog-2198942534740642384.post-17728052499283897982015-07-02T10:55:35.432-07:002015-07-02T10:55:35.432-07:00Very useful!Very useful!Unknownhttps://www.blogger.com/profile/12107442801841553114noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-17821112114820457482014-12-09T12:15:41.557-08:002014-12-09T12:15:41.557-08:00Yes, this can certainly happen.Yes, this can certainly happen.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-55132294190991327922014-12-09T12:14:55.521-08:002014-12-09T12:14:55.521-08:00Is it possible for the sign of the regression coef...Is it possible for the sign of the regression coefficient to change from "+" in OLS to "-" in 2SLS? Or does it imply that there is an issue with the specification?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-65810531952849952772014-05-24T10:38:20.648-07:002014-05-24T10:38:20.648-07:00Re-specify the model. e.g., is the functional for...Re-specify the model. e.g., is the functional form appropriate? Does the model pass the usual specification tests? As a last resort you can constrain the coefficient to be of the desired sign.Dave Gileshttps://www.blogger.com/profile/05389606956062019445noreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-42627473993287235562014-05-24T06:18:45.317-07:002014-05-24T06:18:45.317-07:00so what we can do if we had the wrong sign, while ...so what we can do if we had the wrong sign, while the variable is important?Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-2198942534740642384.post-28081643406304985072013-05-03T13:59:20.989-07:002013-05-03T13:59:20.989-07:00Interesting! It makes sense, but I didn't know...Interesting! It makes sense, but I didn't know it was so simple.<br /><br />Of removing the other variable CAN make the test go from "significant" to "insignificant," which unfortunately is all anybody seems to care about.edhttps://www.blogger.com/profile/01150091053740909530noreply@blogger.com