```Date: Tue, 29 Jul 2008 11:05:37 -0500 Reply-To: "Swank, Paul R" Sender: "SPSSX(r) Discussion" From: "Swank, Paul R" Subject: Re: Multiple Regression Models Comments: To: Arthur Kramer In-Reply-To: <01MXQ0CCSDE69YX5WW@NJCU.edu> Content-Type: text/plain; charset="us-ascii" If I have a significant variable in a model, I try to understand why it's there before I discard it just because it doesn't fit my theory. Sometimes serendipity happens. Paul R. Swank, Ph.D. Professor and Director of Research Children's Learning Institute University of Texas Health Science Center - Houston -----Original Message----- From: Arthur Kramer [mailto:akramer@NJCU.edu] Sent: Tuesday, July 29, 2008 7:56 AM To: Swank, Paul R; SPSSX-L@LISTSERV.UGA.EDU Subject: RE: Multiple Regression Models I think I would look at the "R^2 change" of this variable in the complete model, and then consider the cost/benefit of keeping this model in regard to my theory and what I would be using the model to support. Arthur Kramer, Ph.D. Director of Institutional Research New Jersey City University Phone: 201-200-3073 Fax: 201-200-3288 "...believe half of what you see and none of what you hear." N.Whitfield and B.Strong -----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Swank, Paul R Sent: Monday, July 28, 2008 5:42 PM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Re: Multiple Regression Models The question is, is the increase significant. If so then there are at least two possible reasons. One is power. If the other variables in the model account for a large proportion of the DVs variance, then this variable may become significant because of the increased sensitivity of the F test when the overall R^2 is larger. The second possibility is a suppressor variable. That is, this variable is unrelated to the DV but is related to another IV and suppresses superfluous variance in the other IV and making the model seem stronger. Paul R. Swank, Ph.D. Professor and Director of Research Children's Learning Institute University of Texas Health Science Center - Houston -----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of jimjohn Sent: Monday, July 28, 2008 3:21 PM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Multiple Regression Models when you guys are building multiple regression models, if an independent variable on its own is not significant, but in a regression model with many other independent variables, adding this variable increases the strength of the model (higher Adjusted R^2). In this case, would you recommend keeping the model with the variable that on its own is not significant. Or, would you only consider variables that have significatn pairwise correlations with the dependent variable. Thanks! -- View this message in context: http://www.nabble.com/Multiple-Regression-Models-tp18698762p18698762.htm l Sent from the SPSSX Discussion mailing list archive at Nabble.com. ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ===================== To manage your subscription to SPSSX-L, send a message to LISTSERV@LISTSERV.UGA.EDU (not to SPSSX-L), with no body text except the command. To leave the list, send the command SIGNOFF SPSSX-L For a list of commands to manage subscriptions, send the command INFO REFCARD ```

Back to: Top of message | Previous page | Main SPSSX-L page