Can PROCESS Macro help estimate direct effects of two IVs on one DV with a moderator?

[sennettleung]

Hi! I have proposed this identity model but I am unsure as to how I can test:

the relationship between identity compatibility (which consists of two key
predictors) and well-being (life satisfaction and burnout to be tested
separately) with two moderator variables (independent and interdependent
self-construals to be tested separately as well)
<http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg>
the relationship between identity continuity (which consists of 3 key
predictors) and well-being (life satisfaction and burnout to be tested
separately) with two moderator variables (independent and interdependent
self-construals to be tested separately as well)

on PROCESS macro and/or SPSS?

Any help is much appreciated, thank you so much.



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Rich Ulrich]

If you want to know about PROCESS, you might need to read Hayes's book.

I don't know if it takes two simultaneous predictors. In any case, I personally

am not a fan of packaged analyses where I have to wonder about what the

coefficients, etc., mean, so I would insist on studying the book if I were to

use PROCESS.  Googling shows me examples with z-transforms, which does

simplify certain interpretations.

In terms of analyses, the basic problem is tackled by performing various

multiple regressions, and looking at various results, and drawing logical

conclusions.

You have two predictors; use ENTER=  to put them in after using ENTER=

to put in the moderators of interest.  You are referring to one outcome at

a time. Repeat for your various models.

--

Rich Ulrich

---

From: SPSSX(r) Discussion <[hidden email]> on behalf of sennettleung <[hidden email]>
Sent: Friday, July 13, 2018 10:38:00 AM
To: [hidden email]
Subject: Can PROCESS Macro help estimate direct effects of two IVs on one DV with a moderator?

 

Hi! I have proposed this identity model but I am unsure as to how I can test:

the relationship between identity compatibility (which consists of two key
predictors) and well-being (life satisfaction and burnout to be tested
separately) with two moderator variables (independent and interdependent
self-construals to be tested separately as well)
<http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg>
<http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg>
the relationship between identity continuity (which consists of 3 key
predictors) and well-being (life satisfaction and burnout to be tested
separately) with two moderator variables (independent and interdependent
self-construals to be tested separately as well)

on PROCESS macro and/or SPSS?

Any help is much appreciated, thank you so much.



--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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

[Bruce Weaver]

I agree with Rich.  Given that you are looking at interactions only, and not
looking at mediation effects, there is no need, and as far as I can tell, no
real advantage for PROCESS.  If all explanatory variables are quantitative,
you can use either REGRESSION or UNIANOVA to estimate your models.  If any
variables are categorical, you may find it more convenient to use UNIANOVA,
as it allows you to enter categorical variables as "fixed factors" (where
REGRESSION would require you to compute k-1 indicators for each categorical
variable, and more indicators for product terms if there were interactions).
One advantage of using UNIANOVA is that you can use multiple EMMEANS
sub-commands to get fitted values of Y at selected combinations of the
explanatory variables.  Plotting these can help you to visualize the nature
of the interaction(s).  

HTH.


Rich Ulrich wrote

> If you want to know about PROCESS, you might need to read Hayes's book.
>
> I don't know if it takes two simultaneous predictors. In any case, I
> personally
>
> am not a fan of packaged analyses where I have to wonder about what the
>
> coefficients, etc., mean, so I would insist on studying the book if I were
> to
>
> use PROCESS.  Googling shows me examples with z-transforms, which does
>
> simplify certain interpretations.
>
>
> In terms of analyses, the basic problem is tackled by performing various
>
> multiple regressions, and looking at various results, and drawing logical
>
> conclusions.
>
>
> You have two predictors; use ENTER=  to put them in after using ENTER=
>
> to put in the moderators of interest.  You are referring to one outcome at
>
> a time. Repeat for your various models.
>
>
> --
>
> Rich Ulrich
>
> ________________________________
> From: SPSSX(r) Discussion &lt;

> SPSSX-L@.UGA

> &gt; on behalf of sennettleung &lt;

> sennett_0214@

> &gt;
> Sent: Friday, July 13, 2018 10:38:00 AM
> To:

> SPSSX-L@.UGA

> Subject: Can PROCESS Macro help estimate direct effects of two IVs on one
> DV with a moderator?
>
> Hi! I have proposed this identity model but I am unsure as to how I can
> test:
>
> the relationship between identity compatibility (which consists of two key
> predictors) and well-being (life satisfaction and burnout to be tested
> separately) with two moderator variables (independent and interdependent
> self-construals to be tested separately as well)
> &lt;http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg&gt;
> &lt;http://spssx-discussion.1045642.n5.nabble.com/file/t341477/IMG_8752_%282%29-min.jpg&gt;
> the relationship between identity continuity (which consists of 3 key
> predictors) and well-being (life satisfaction and burnout to be tested
> separately) with two moderator variables (independent and interdependent
> self-construals to be tested separately as well)
>
> on PROCESS macro and/or SPSS?
>
> Any help is much appreciated, thank you so much.
>
>
>
> --
> Sent from: http://spssx-discussion.1045642.n5.nabble.com/
>
> =====================
> To manage your subscription to SPSSX-L, send a message to

> LISTSERV@.UGA

>  (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@.UGA

>  (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





-----
--
Bruce Weaver
[hidden email]
http://sites.google.com/a/lakeheadu.ca/bweaver/

"When all else fails, RTFM."

NOTE: My Hotmail account is not monitored regularly.
To send me an e-mail, please use the address shown above.

--
Sent from: http://spssx-discussion.1045642.n5.nabble.com/

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