Two-way interaction (product-term) model to plot the interactive graph

[David Lin-4]

I have a two-way interaction (product-term) model, and I want to plot the
interactive graph. I tried using multiple regression analysis to estimate
the conditional and interaction effects of a set of two measures (dev,cer)
on SC , without using ANOVA-based procedures which required the arbitrary
dichotomization of independent variables. All variables are continuous
variables.

With the regression equation SC=(a+b1E1+b2E2+b3E1E2),
that can be rewritten as SC=(a+b2E2)+(b1+b3E2)E1.

I want to plot the interactive graph by making E2=-1, 0, and 1,
respectively. How can I do this by SPSS syntax?


/* two-way interaction (product-term) model
COMPUTE devcer = dev*cer.
EXE.
REGRESSION
/STATISTICS COEFF OUTS R ANOVA COLLIN TOL ZPP CHA
/CRITERIA=PIN(.05) POUT(.10)
/NOORIGIN
/DEPENDENT sc
/METHOD=ENTER dev cer
/METHOD=ENTER devcer.

/* My problem here. This is ANOVA-based procedure, not what I want.
IGRAPH
/VIEWNAME='Scatterplot'
/X1 = VAR(dev) TYPE = SCALE
/X2 = VAR(cer) TYPE = SCALE
/Y = VAR(sc) TYPE = SCALE
/COLOR = VAR(group) TYPE = CATEGORICAL
/COORDINATE = VERTICAL
/FITLINE METHOD = REGRESSION LINEAR LINE = MEFFECT SPIKE=OFF
/X1LENGTH=3.0
/YLENGTH=3.0
/X2LENGTH=3.0
/CHARTLOOK='NONE'
/CATORDER VAR(group) (ASCENDING VALUES OMITEMPTY)
/SCATTER COINCIDENT =NONE.
EXE.

=====================
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

[Marta Garcia-Granero]

Hi David

Not an SPSS solution, but you can use ITALASSI, a freeware program that
plots two-way interactions between continuous predictors. It can read
SPSS files directly.

See a short description of the program features:
http://www.provalisresearch.com/ITALASSI/ITALASSI.php

You can download it from here:
http://www.provalisresearch.com/ITALASSI/ITALdowload.html

HTH,
Marta García-Granero

> I have a two-way interaction (product-term) model, and I want to plot the
> interactive graph. I tried using multiple regression analysis to estimate
> the conditional and interaction effects of a set of two measures (dev,cer)
> on SC , without using ANOVA-based procedures which required the arbitrary
> dichotomization of independent variables. All variables are continuous
> variables.
>
> With the regression equation SC=(a+b1E1+b2E2+b3E1E2),
> that can be rewritten as SC=(a+b2E2)+(b1+b3E2)E1.
>
> I want to plot the interactive graph by making E2=-1, 0, and 1,
> respectively. How can I do this by SPSS syntax?
>
>
> /* two-way interaction (product-term) model
> COMPUTE devcer = dev*cer.
> EXE.
> REGRESSION
> /STATISTICS COEFF OUTS R ANOVA COLLIN TOL ZPP CHA
> /CRITERIA=PIN(.05) POUT(.10)
> /NOORIGIN
> /DEPENDENT sc
> /METHOD=ENTER dev cer
> /METHOD=ENTER devcer.
>
> /* My problem here. This is ANOVA-based procedure, not what I want.
> IGRAPH
> /VIEWNAME='Scatterplot'
> /X1 = VAR(dev) TYPE = SCALE
> /X2 = VAR(cer) TYPE = SCALE
> /Y = VAR(sc) TYPE = SCALE
> /COLOR = VAR(group) TYPE = CATEGORICAL
> /COORDINATE = VERTICAL
> /FITLINE METHOD = REGRESSION LINEAR LINE = MEFFECT SPIKE=OFF
> /X1LENGTH=3.0
> /YLENGTH=3.0
> /X2LENGTH=3.0
> /CHARTLOOK='NONE'
> /CATORDER VAR(group) (ASCENDING VALUES OMITEMPTY)
> /SCATTER COINCIDENT =NONE.
> EXE.
>
> =====================
> 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
>
>

=====================
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