LISTSERV at the University of Georgia
Menubar Imagemap
Home Browse Manage Request Manuals Register
Previous messageNext messagePrevious in topicNext in topicPrevious by same authorNext by same authorPrevious page (December 2011)Back to main SPSSX-L pageJoin or leave SPSSX-L (or change settings)ReplyPost a new messageSearchProportional fontNon-proportional font
Date:         Mon, 19 Dec 2011 17:07:44 +0000
Reply-To:     "Poes, Matthew Joseph" <mpoes@UILLINOIS.EDU>
Sender:       "SPSSX(r) Discussion" <SPSSX-L@LISTSERV.UGA.EDU>
From:         "Poes, Matthew Joseph" <mpoes@UILLINOIS.EDU>
Subject:      Re: Can I compare two within-subjects' dependent variables while
              holding co-variates?
Comments: To: Gene Maguin <emaguin@buffalo.edu>
In-Reply-To:  <004701ccbe6d$52825e50$f7871af0$@buffalo.edu>
Content-Type: text/plain; charset="us-ascii"

Having had extensive training/education in the use of MLM/HLM (Done through mixed modeling in SPSS and SAS), and having used it extensively in my evaluation work, I get a bit nervous when I see the potential for people to jump into this form of modeling based solely on self-teaching. The SPSS documentation is wholly inadequate in my opinion, and there are some problems with the Mixed models used in SPSS, which requires an understanding of how they work to appropriately interpret them.

Common pitfalls are a misunderstanding of how to interpret the coefficients, and to over-estimate the model, without understanding the impact that is having.

I do agree with Gene that a mixed model approach (More commonly known in the literature as multi-level modeling (MLM) or Hierarchical Linear Modeling (HLM) is the way to go, but for different reasons. The data you have has an event nested within a person, and a person nested within a couple. Thought of another way (and this is modeled differently), you might say that you have a person, nested within a couple, nested within an event. The latter would mean that you believe that responses to a particular even are more likely to be similar to each other than they would between events. This can thus be explicitly modeled.

In terms of how to "separate" the duration from the model, you have to make a decision. Do you want to model the events separately, which will completely eliminate the effect of one event on the other, or do you want to include the other event in the modeling of the latter. If the Latter is the case, then you can do this with interaction terms of duration and event. However, you still would have the effect of the other event and it's duration modeled in the overall tests of effects.

That brings up a different issue. Are you modeling two events as separate outcome variables? Multiple dependent variables cannot be modeled in the mixed method program in SPSS or SAS, and it's actually been argued that they shouldn't be anyway. Also, SPSS generally only has linear mixed models, which may be a limiting factor for your analysis. In 19 they did add a mixed GLM approach; however I've seen little documentation on it, and only just begun using it and comparing it with the HLM package, which I trust much more.

Matthew J Poes Research Data Specialist Center for Prevention Research and Development University of Illinois

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of Gene Maguin Sent: Monday, December 19, 2011 10:43 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Re: Can I compare two within-subjects' dependent variables while holding co-variates?

NHR,

You are going to have to restructure your data and use Mixed. Look through the documentation in the syntax reference. I think what you are describing is explicitly mentioned, although I am not sure.

Gene Maguin

-----Original Message----- From: SPSSX(r) Discussion [mailto:SPSSX-L@LISTSERV.UGA.EDU] On Behalf Of NHR Sent: Monday, December 19, 2011 11:22 AM To: SPSSX-L@LISTSERV.UGA.EDU Subject: Re: Can I compare two within-subjects' dependent variables while holding co-variates?

Hi, Thanks a lot for your replay.

My dependent variables are PTSD measures. I have 3 sub-scales and one total measure. These 4 measures refer to two different events, so each participant has 8 measures (4 for every event - 4 couples).

I wanted to compare, for example the degree of avoidance symptoms between event 1 and 2, and so on, regarding each couple. Each participant has 2 marks/grads, for each measure - one for every event. But, than I wand to hold as co-variates the time duration that passed since each event. I wanted to do that cause there is a significant difference between the time duration that passed between the two events. While doing an ANOVA or MANOVA, the two co-variates (duration since each event) influence the two dependent variables, and that is no good. I thinks there is no point in holding as co-variate the duration since event 1 while looking at the symptomolgy after event 2. So, I was wondering how can I compare my dependent variables in a better way?

The participants were all volunteers, who agreed to participate in a web-survey.

I hope my answers are clear. Please let me know if not, and I'll be happy to answer.

Thanks a lot, I really appreciate it, NHR

-- View this message in context: http://spssx-discussion.1045642.n5.nabble.com/Can-I-compare-two-within-subje cts-dependent-variables-while-holding-co-variates-tp5085591p5086603.html 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