Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Predict command

From   "Mary E. Mackesy-Amiti" <>
Subject   Re: st: Predict command
Date   Wed, 21 Apr 2010 10:32:18 -0500

If something that is significant in the full model becomes non-significant in the reduced model that would suggest that that "extra" variable somehow modifies the relationship between your dependent and independent variables. Instead of ignoring it you should take a closer look and figure out what is going on. Look up "suppressor variables".

Rasika Raghavan wrote:

I want to fit values to a regression model based on only the
significant variables. I am aware that I have the option of using only
the significant varibales in the model , re running the regression and
then using the predict command. But it is likely that these
significant variables might become insignificant. Hence I do not want
to tamper with the model but just predict based on the significant
variables.Another option would be to output the coefficients and then
calculate the predicted values through the relevant multiplication and
additions. Is there but a direct way?

*   For searches and help try:

Mary Ellen Mackesy-Amiti, Ph.D.
Research Assistant Professor
Community Outreach Intervention Projects (COIP)
School of Public Health m/c 923
Division of Epidemiology and Biostatistics
University of Illinois at Chicago
ph. 312-355-4892
fax: 312-996-1450

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index