Mixed or GLM procedures
Locked
 
|
I am analysing fixed effects for country data over a number of years. I have 3 independent variables for which I also wish to generate all interaction terms aside from direct effects terms.
I attempted both MiXED and GLM repeated procedures on SPSS 18 choosing the factorial option (I also tried the interaction and 3-way option) but in the output I only see direct effects coefficients and not the 2-way and three-way terms and co-efficients. Will appreciate assistance around this.
Also does the MIXED procure automatically generate dummies for the countries or do I have to do this using RECODE and input these as factors.?
Colin This email and all contents are subject to the following disclaimer: http://disclaimer.uj.ac.za |
|
|
Colin If you specify country as a factor (not a covariate!), MIXED will generate dummmies automatically. You can specify the terms in your model including interactions using the ‘Fixed’ dialog. On the LH side of the dialog box, highlight the variables you want in the Factors and Covariates box, select Factorial in the dropdown menu and click Add. The main effects and interactions should then appear in the RH box titled ‘Model’ . If that is not working I haven’t a cluse what is going on. However, you could paste the syntax, and examine the result, and if necessary, respecify the main effects and interactions you want under the /FIXED portion of the MIXED command. Garry Gelade From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Reddy, Colin I am analysing fixed effects for country data over a number of years. I have 3 independent variables for which I also wish to generate all interaction terms aside from direct effects terms. I attempted both MiXED and GLM repeated procedures on SPSS 18 choosing the factorial option (I also tried the interaction and 3-way option) but in the output I only see direct effects coefficients and not the 2-way and three-way terms and co-efficients. Will appreciate assistance around this. Also does the MIXED procure automatically generate dummies for the countries or do I have to do this using RECODE and input these as factors.? Colin
|
|
|
In reply to this post by Reddy, Colin
It is not completely clear what you're saying below.
Consider:
(1) It sounds like you're using the menuing system to
generate
the syntax for the two procedures. I recommend that you
take
the syntax that is generated and rewrite the bits that fit
into your
overall design.
(2) You say you have three "independent
variables". Are these
variables all between-subjects factors? If so,
then the syntax for
GLM should look like this:
(Note: DV=Dependent Variable, A=1st ind var, B=2nd ind
var,
C=3 ind var)
GLM DV by A B C/
print=descriptives homogeneity/
Design=A,B,C,A*B,A*C,B*C,A*B*C.
You can add additional subcommands for things but the
above
specifies the design you want to analyze, at least from what
I
can glean from what you say below.
(3) You should be able to specify the above design
through
the menu system, so I wonder did you get any error
messages?
Additionally, what is the total number of units of analysis
and
how serious is the imbalance in sample sizes, especially
in
cells of the three-way interaction? If you specified the
correct
design in the menu system and you didn't get 2-way and
3-way
interactions, then maybe they could not be calculated (any
cells
with N=0?).
I can't help you with the Mixed procedure but there are
others
on the list who can.
-Mike Palij
New York University
|
|
|
In reply to this post by Reddy, Colin
"data over a number of years" by you are "analysing fixed effects".
What is the role of "years" in your data, ... hypotheses, ... design? -- Rich Ulrich Date: Tue, 19 Jul 2011 13:47:11 +0000 From: [hidden email] Subject: Mixed or GLM procedures To: [hidden email]
I am analysing fixed effects for country data over a number of years. I have 3 independent variables for which I also wish to generate all interaction terms aside from direct effects terms. I attempted both MiXED and GLM repeated procedures on SPSS 18 choosing the factorial option (I also tried the interaction and 3-way option) but in the output I only see direct effects coefficients and not the 2-way and three-way terms and co-efficients. Will appreciate assistance around this. Also does the MIXED procure automatically generate dummies for the countries or do I have to do this using RECODE and input these as factors.? Colin |
|
|
In reply to this post by Reddy, Colin
Colin,
It would help if you posted (1) your research questions, (2) study design, (3) an illustration of your dataset, and (4) syntax. Ryan On Tue, Jul 19, 2011 at 9:47 AM, Reddy, Colin <[hidden email]> wrote: > I am analysing fixed effects for country data over a number of years. I have > 3 independent variables for which I also wish to generate all interaction > terms aside from direct effects terms. > > > > I attempted both MiXED and GLM repeated procedures on SPSS 18 choosing the > factorial option (I also tried the interaction and 3-way option) but in the > output I only see direct effects coefficients and not the 2-way and > three-way terms and co-efficients. Will appreciate assistance around this. > > > > Also does the MIXED procure automatically generate dummies for the countries > or do I have to do this using RECODE and input these as factors.? > > > > Colin > > ________________________________ > This email and all contents are subject to the following disclaimer: > > http://disclaimer.uj.ac.za > ===================== To manage your subscription to SPSSX-L, send a message to [hidden email] (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 |
| Free forum by Nabble | Edit this page |