moderation (not exactly spss-y)

[Nika Yugay]

Hello dear list!

I have a question about moderation - how can i explain the following:

i get a significant moderation effect (did by regression); however
single-slope analysis shows that neither of the slopes is significantly
different... help help please please!

thanks a lot,

Nika

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

[Hector Maletta]

         You mean the original slope X on Y gets less steep when you control
for Z, but the two lines (for high and low Z) are parallel, i.e. with the
same slope. If this is the case, it would indicate moderation without
interaction. Z moderates the effect of X, but does not interact with X: its
moderating effect is the same at all levels of X.

         Hector





         -----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Nika Yugay
Sent: 24 October 2007 10:17
To: [hidden email]
Subject: moderation (not exactly spss-y)

         Hello dear list!

         I have a question about moderation - how can i explain the
following:

         i get a significant moderation effect (did by regression); however
         single-slope analysis shows that neither of the slopes is
significantly
         different... help help please please!

         thanks a lot,

         Nika

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

[Anila Qureshi]

Hi all,
I am working with SPSS 15. I am trying to get a 20% random sample
proprotionate to each region (my file has 11 regions). I know I can do it by
filtering out region and running random sample query on each.  I am just
curious to know is there any single syntax command that does it for all
region.

 Thanks

>
>
>
>
>
>         -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> Nika Yugay
> Sent: 24 October 2007 10:17
> To: [hidden email]
> Subject: moderation (not exactly spss-y)
>
>         Hello dear list!
>
>         I have a question about moderation - how can i explain the
> following:
>
>         i get a significant moderation effect (did by regression); however
>         single-slope analysis shows that neither of the slopes is
> significantly
>         different... help help please please!
>
>         thanks a lot,
>
>         Nika
>
>         =====================
>         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
>

--
Anila Qureshi

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

[William B. Ware]

Nika,

Neither slope is different from "what?"  It sounds as though you looked at
each one separately... the test of the interaction is comparing them to
one another...

wbw

__________________________________________________________________________
William B. Ware, Professor                         Educational Psychology,
CB# 3500                                       Measurement, and Evaluation
University of North Carolina                         PHONE  [phone]
Chapel Hill, NC      27599-3500                      FAX:   [phone]
Office:  118 Peabody Hall                            EMAIL: [hidden email]
Adjunct Professor                                    School of Social Work
__________________________________________________________________________


On Wed, 24 Oct 2007, Nika Yugay wrote:

> Hello dear list!
>
> I have a question about moderation - how can i explain the following:
>
> i get a significant moderation effect (did by regression); however
> single-slope analysis shows that neither of the slopes is significantly
> different... help help please please!
>
> thanks a lot,
>
> Nika
>
> =====================
> 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

[Richard Ristow]

(By the by, it's much the best to use a new and descriptive subject
head, when you're asking a new question. I'll bet a lot of people who
might be helpful, won't have read yours.)

At 10:30 AM 10/24/2007, Anila Qureshi wrote:

>I am working with SPSS 15. I am trying to get a 20% random sample
>proprotionate to each region (my file has 11 regions).

First, there's always

That's very easy. It'll select each case with probability 0.2. You'll
get *approximate* 20% samples overall, and by region. Since mild
over-sampling is usually OK, you can reduce the risk of under-sampling
by increasing the size a bit:

SAMPLE 0.25.

Second, for *exact* sampling by region, I'd hard-code the "k/n" method.
In this code (not tested, but based on tested code) I round the 20%
sample size up to the next integer. This does *not* select cases, but
marks them as "InSample" or not. Select or filter later, as desired.

NUMERIC    InSample (F2).
VAR LABELS InSample 'Selected for 20% sample by region?'.
VAL LABELS InSample 1 'Selected'  0 'Not selected'.

SORT CASES BY Region.

AGGREGATE OUTFILE=*
    /BREAK = Region
    /RegionTot 'Total cases in region' = NU.

DO IF       MISSING(LAG(Region))
          OR Region NE LAG(Region)).
.   COMPUTE #N = RegionTot.
.   COMPUTE #K = RND((0.20*RegionTot+0.5).
END IF.

compute InSample = RV.BERNOULLI(#K/#N).
compute #K =  #K - InSample.
compute #N = #N - 1.

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

[Anila Qureshi]

Thanks a lot Ristow. I'll work on it.
I am pretty new to the list and thats why I missed a descriptive subject.
Sorry for that!
This listserv is an excellent learning source; I love it.
Thanks again!

On 10/24/07, Richard Ristow <[hidden email]> wrote:

>
> (By the by, it's much the best to use a new and descriptive subject
> head, when you're asking a new question. I'll bet a lot of people who
> might be helpful, won't have read yours.)
>
> At 10:30 AM 10/24/2007, Anila Qureshi wrote:
>
> >I am working with SPSS 15. I am trying to get a 20% random sample
> >proprotionate to each region (my file has 11 regions).
>
> First, there's always
>
> That's very easy. It'll select each case with probability 0.2. You'll
> get *approximate* 20% samples overall, and by region. Since mild
> over-sampling is usually OK, you can reduce the risk of under-sampling
> by increasing the size a bit:
>
> SAMPLE 0.25.
>
> Second, for *exact* sampling by region, I'd hard-code the "k/n" method.
> In this code (not tested, but based on tested code) I round the 20%
> sample size up to the next integer. This does *not* select cases, but
> marks them as "InSample" or not. Select or filter later, as desired.
>
> NUMERIC    InSample (F2).
> VAR LABELS InSample 'Selected for 20% sample by region?'.
> VAL LABELS InSample 1 'Selected'  0 'Not selected'.
>
> SORT CASES BY Region.
>
> AGGREGATE OUTFILE=*
>    /BREAK = Region
>    /RegionTot 'Total cases in region' = NU.
>
> DO IF       MISSING(LAG(Region))
>          OR Region NE LAG(Region)).
> .   COMPUTE #N = RegionTot.
> .   COMPUTE #K = RND((0.20*RegionTot+0.5).
> END IF.
>
> compute InSample = RV.BERNOULLI(#K/#N).
> compute #K =  #K - InSample.
> compute #N = #N - 1.
>
>

--
Anila Qureshi

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