One tail test

[juanreynel]

the area below 360 of a randomly normal distributed variable X is 0.494 with mean 400 and 50 ST. Is it possible to shade this area in SPSS, graphically or programmatically?

[Andy W]

I wrote a blog post just the other day showing how one could do this, see http://andrewpwheeler.wordpress.com/2013/03/06/some-random-spss-graph-tips-shading-areas-under-curves-and-using-dodging-in-binned-dot-plots/. I hope it is somewhat obvious how to adapt it to new situations.

It is obvious from the graph your numbers are off, whether your talking about the marginal density or the cumulative density. Or the know the direct easy way in calculations shown after the graph command.

*****************************************.
input program.
loop #i = 1 to 2000.
compute X = (#i - 1000)/5 + 400.
compute PDF = PDF.NORMAL(X,400,50).
compute CDF = CDF.NORMAL(X,400,50).
end case.
end loop.
end file.
end input program.
dataset name sim.
exe.

formats PDF X (F2.1).

*Mark off different areas.
compute tails = 0.
if X < 360 tails = 1.

*Area with particular locations highlighted.
GGRAPH
  /GRAPHDATASET NAME="graphdataset" VARIABLES=X PDF tails
  /GRAPHSPEC SOURCE=INLINE.
BEGIN GPL
 SOURCE: s=userSource(id("graphdataset"))
 DATA: X=col(source(s), name("X"))
 DATA: PDF=col(source(s), name("PDF"))
 DATA: tails=col(source(s), name("tails"), unit.category())
 SCALE: cat(aesthetic(aesthetic.color.interior), map(("0",color.white),("1",color.grey),("2",color.grey)))
 SCALE: cat(aesthetic(aesthetic.transparency.interior), map(("0",transparency."1"),("1",transparency."0"),("2",transparency."0")))
 GUIDE: axis(dim(1), label("X"))
 GUIDE: axis(dim(2), label("Prob. Dens."))
 GUIDE: legend(aesthetic(aesthetic.color.interior), null())
 GUIDE: legend(aesthetic(aesthetic.transparency.interior), null())
 ELEMENT: area(position(X*PDF), color.interior(tails), transparency.interior(tails))
END GPL.

*What is the actual area?.
*Do you mean PDF or CDF?.
temporary.
select if X = 360.
freq var PDF CDF.
exe.
*Neither are ".494".
*****************************************.

[Bruce Weaver]

That looks like a useful blog post.  Thanks Andy.

Andy W wrote

I wrote a blog post just the other day showing how one could do this, see http://andrewpwheeler.wordpress.com/2013/03/06/some-random-spss-graph-tips-shading-areas-under-curves-and-using-dodging-in-binned-dot-plots/. I hope it is somewhat obvious how to adapt it to new situations.

It is obvious from the graph your numbers are off, whether your talking about the marginal density or the cumulative density. Or the know the direct easy way in calculations shown after the graph command.

*****************************************.
input program.
loop #i = 1 to 2000.
compute X = (#i - 1000)/5 + 400.
compute PDF = PDF.NORMAL(X,400,50).
compute CDF = CDF.NORMAL(X,400,50).
end case.
end loop.
end file.
end input program.
dataset name sim.
exe.

formats PDF X (F2.1).

*Mark off different areas.
compute tails = 0.
if X < 360 tails = 1.

*Area with particular locations highlighted.
GGRAPH
  /GRAPHDATASET NAME="graphdataset" VARIABLES=X PDF tails
  /GRAPHSPEC SOURCE=INLINE.
BEGIN GPL
 SOURCE: s=userSource(id("graphdataset"))
 DATA: X=col(source(s), name("X"))
 DATA: PDF=col(source(s), name("PDF"))
 DATA: tails=col(source(s), name("tails"), unit.category())
 SCALE: cat(aesthetic(aesthetic.color.interior), map(("0",color.white),("1",color.grey),("2",color.grey)))
 SCALE: cat(aesthetic(aesthetic.transparency.interior), map(("0",transparency."1"),("1",transparency."0"),("2",transparency."0")))
 GUIDE: axis(dim(1), label("X"))
 GUIDE: axis(dim(2), label("Prob. Dens."))
 GUIDE: legend(aesthetic(aesthetic.color.interior), null())
 GUIDE: legend(aesthetic(aesthetic.transparency.interior), null())
 ELEMENT: area(position(X*PDF), color.interior(tails), transparency.interior(tails))
END GPL.

*What is the actual area?.
*Do you mean PDF or CDF?.
temporary.
select if X = 360.
freq var PDF CDF.
exe.
*Neither are ".494".
*****************************************.

[David Marso]

This looks interesting.  
GPL is on my list of 'to conquer' along with python since my graduation to ver 21.
However I am rather dismayed by the following from the FM:
This little paragraph ?? Anyone care to explain why/how this came to pass?
Seems very short sighted and somewhat limiting!
So If I want to make my code general wrt GPL I need to use python?
Seriously? What are they putting in the water in Chicago?
---
Scope and Limitations
 GPL blocks cannot be nested within GPL blocks.
 GPL blocks cannot be contained within DEFINE-!ENDDEFINE macro definitions.
 GPL blocks can be contained in command syntax files run via the INSERT command, with
the default SYNTAX=INTERACTIVE setting.
 GPL blocks cannot be contained within command syntax files run via the INCLUDE command.

[Jon K Peck]

The GPL language specification, which is cross product, and the macro parser conflict and can't be used together.  Not our preference, but  nothing we could do about it.  However it is super easy to make the equivalent of a macro with parameter substitution with a few lines of Python code.  Or, with a few more lines, to make an extension command.


Jon Peck (no "h") aka Kim
Senior Software Engineer, IBM
[hidden email]
phone: [phone]




From:        David Marso <[hidden email]>
To:        [hidden email],
Date:        03/13/2013 05:20 PM
Subject:        Re: [SPSSX-L] One tail test
Sent by:        "SPSSX(r) Discussion" <[hidden email]>

---

This looks interesting.
GPL is on my list of 'to conquer' along with python since my graduation to
ver 21.
However I am rather dismayed by the following from the FM:
This little paragraph ?? Anyone care to explain why/how this came to pass?
Seems very short sighted and somewhat limiting!
So If I want to make my code general wrt GPL I need to use python?
Seriously? What are they putting in the water in Chicago?
---
Scope and Limitations
�� GPL blocks cannot be nested within GPL blocks.
�� GPL blocks *cannot be contained within DEFINE-!ENDDEFINE* macro
definitions.
�� GPL blocks can be contained in command syntax files run via the INSERT
command, with
the default SYNTAX=INTERACTIVE setting.
�� GPL blocks cannot be contained within command syntax files run via the
INCLUDE command.



-----
Please reply to the list and not to my personal email.
Those desiring my consulting or training services please feel free to email me.
--
View this message in context: http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718634.html
Sent from the SPSSX Discussion mailing list archive at 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

[Andy W]

Marta Garcia-Granero actually has an example of placing inline GPL code within a macro, but it is difficult and annoying for sure, http://gjyp.nl/marta/Kalbfleisch&Prentice95%25CI.SPS

The last time I tried this the GPL code won't print when you run the macro, so debugging is very difficult (and I gave up). Basically what I do now is have boiler-plate GPL code for what I want to do with anything complicated (e.g. like this, http://andrewpwheeler.wordpress.com/2013/02/28/interval-graph-for-viz-temporal-overlap-in-crime-events/).

My current job I have a version pre-python (or at least the python installation for it has long since been unavailable from online), but it is on my to-do list to better integrate python as well (and make real GUI dialogs for some of my scripts). A few examples using GGRAPH are floating around, see one blog post by Jon (https://www.ibm.com/developerworks/mydeveloperworks/blogs/ab16c38e-2f7b-4912-a47e-85682d124d32/entry/how_can_i_parameterize_ggraph_and_gpl_code6?lang=en) and this answer on stackoverflow (http://stackoverflow.com/a/13825642/604456). I believe the person who gave that example runs this site, Python for SPSS, http://www.pythonforspss.org/index.html, which may be of interest to the general community here.

I see in some googling Raynald has a very tiny sampling of python examples on his site, http://www.spsstools.net/python.htm.

[David Marso]

Hi Andy,
Thanks for those links.  I will have to study these.  As soon as I get
done rolling a few rocks up the hill I plan to bust out python and
become a snakehead.  Also GPL and other fun stuff.
Hell, maybe I'll even develop some scripts to beat up on my new
nemesis the Model Viewer ;-))
---
You may have missed this classic exchange.

http://spssx-discussion.1045642.n5.nabble.com/The-agony-and-the-ecstacy-Development-of-complex-MATRIX-programs-td5716810.html#a5716842
---

I was actually going to ask if you would be willing to host a 'guest
blogger' on your wordpress site  Particularly targeted at putting some
meat on the technique I introduce half way down in this thread (hence
I tracked it down and what do you know?).
Interesting convergence there:
Odd that the code you cite from Marta is actually part of the above thread ;-)
In particular, the recent thread on rolling your own bootstrap could
be incorporated into the article and it might be a nice collaboration
(I would say also roll a bunch of stats as well).

At the moment I am in the midst of my 2000 line MATRIX MACRO program
(yeah, I'm still (arguably) sane) and at some point I will need to
generate some fairly complex graphics so this might be a little in the
future or between nervous breakdowns ;-).
David
--
On Thu, Mar 14, 2013 at 3:37 PM, Andy W [via SPSSX Discussion]
<[hidden email]> wrote:

> Marta Garcia-Granero actually has an example of placing inline GPL code
> within a macro, but it is difficult and annoying for sure,
> http://gjyp.nl/marta/Kalbfleisch&Prentice95%25CI.SPS
>
> The last time I tried this the GPL code won't print when you run the macro,
> so debugging is very difficult (and I gave up). Basically what I do now is
> have boiler-plate GPL code for what I want to do with anything complicated
> (e.g. like this,
> http://andrewpwheeler.wordpress.com/2013/02/28/interval-graph-for-viz-temporal-overlap-in-crime-events/).
>
> My current job I have a version pre-python (or at least the python
> installation for it has long since been unavailable from online), but it is
> on my to-do list to better integrate python as well (and make real GUI
> dialogs for some of my scripts). A few examples using GGRAPH are floating
> around, see one blog post by Jon
> (https://www.ibm.com/developerworks/mydeveloperworks/blogs/ab16c38e-2f7b-4912-a47e-85682d124d32/entry/how_can_i_parameterize_ggraph_and_gpl_code6?lang=en)
> and this answer on stackoverflow
> (http://stackoverflow.com/a/13825642/604456). I believe the person who gave
> that example runs this site, Python for SPSS,
> http://www.pythonforspss.org/index.html, which may be of interest to the
> general community here.
>
> I see in some googling Raynald has a very tiny sampling of python examples
> on his site, http://www.spsstools.net/python.htm.
>
>
> Andy W
> [hidden email]
> http://andrewpwheeler.wordpress.com/
>
>
> ________________________________
> If you reply to this email, your message will be added to the discussion
> below:
> http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718669.html
> To unsubscribe from One tail test, click here.
> NAML

[Andy W]

I'm totally cool with a guest post, although my readership isn't anything to brag about! Just shoot me an email (not sure how the logistics work in wordpress but I'm sure I can figure it out).

I was going to say something about Ryan's code on estimating Rasch models on my blog as well (sometime). That is a perfect example, instead of making a question linger on the list-serve better to get it out there on a website (and not have good things lost to the wind if Nabble goes away tomorrow). Alot of people have code posted on the web to the same effect (e.g. Marta, Bruce, Raynald), but I enjoy blogging about it in addition to that.

Offer for guest posting on my blog is up to anyone. In the end I would like a SPSS blog roll like there is for R-bloggers, so make your own blogs! (And let me know about it)

Andy

ps - Those who consult, a blog is an excellent way to bring notice to your work. It may be purely coincidental but I got a pretty good gig with some people from the public policy dept. at my University that wouldn't have otherwise known about me.

[David Marso]

I am actually on a project that came word of mouth from another client
who I found from an SPSS listing.  All I had to do was link to my
profile and it was !JOB IS YOURS!.
Actually, Nabble is only one archive (UGA is the original :
http://listserv.uga.edu/cgi-bin/wa?A0=spssx-l&D=0&F=&H=0&O=T&S=&T=1

Unfortunately a ton of stuff somehow got nuked!
It appears to begin in Feb 1996 but that is oh so wrong!
Or maybe comp.soft-sys.stat.spss (which has been spammed to death)
Upon inspection it appears that  is nuked back to 1996 as well.  I
clearly recall posting numerous gems in the early 1990's and was
nearly out of support in 1996.  Nowdays everything I post of value
accumulates on my HD.
Yeah, If it is of any value it can be recreated!
Let's see how the time works in coming weeks.
BTW: Are you on LinkedIn?
David

On Thu, Mar 14, 2013 at 4:29 PM, Andy W [via SPSSX Discussion]
<[hidden email]> wrote:

> I'm totally cool with a guest post, although my readership isn't anything to
> brag about! Just shoot me an email (not sure how the logistics work in
> wordpress but I'm sure I can figure it out).
>
> I was going to say something about Ryan's code on estimating Rasch models on
> my blog as well (sometime). That is a perfect example, instead of making a
> question linger on the list-serve better to get it out there on a website
> (and not have good things lost to the wind if Nabble goes away tomorrow).
> Alot of people have code posted on the web to the same effect (e.g. Marta,
> Bruce, Raynald), but I enjoy blogging about it in addition to that.
>
> Offer for guest posting on my blog is up to anyone. In the end I would like
> a SPSS blog roll like there is for R-bloggers, so make your own blogs! (And
> let me know about it)
>
> Andy
>
> ps - Those who consult, a blog is an excellent way to bring notice to your
> work. It may be purely coincidental but I got a pretty good gig with some
> people from the public policy dept. at my University that wouldn't have
> otherwise known about me.
> Andy W
> [hidden email]
> http://andrewpwheeler.wordpress.com/
>
>
> ________________________________
> If you reply to this email, your message will be added to the discussion
> below:
> http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718673.html
> To unsubscribe from One tail test, click here.
> NAML

[Art Kendall]

IIRC there was a discussion list or newsgroup back in the mid 80's.  Was it on BITNET? It was something later than ARPAnet or DECnet.

It might have been one of the early competitors to L-Soft's LISTSERV.
On 3/14/2013 5:35 PM, David Marso wrote:

I am actually on a project that came word of mouth from another client
who I found from an SPSS listing.  All I had to do was link to my
profile and it was !JOB IS YOURS!.
Actually, Nabble is only one archive (UGA is the original :
http://listserv.uga.edu/cgi-bin/wa?A0=spssx-l&D=0&F=&H=0&O=T&S=&T=1

Unfortunately a ton of stuff somehow got nuked!
It appears to begin in Feb 1996 but that is oh so wrong!
Or maybe comp.soft-sys.stat.spss (which has been spammed to death)
Upon inspection it appears that  is nuked back to 1996 as well.  I
clearly recall posting numerous gems in the early 1990's and was
nearly out of support in 1996.  Nowdays everything I post of value
accumulates on my HD.
Yeah, If it is of any value it can be recreated!
Let's see how the time works in coming weeks.
BTW: Are you on LinkedIn?
David

On Thu, Mar 14, 2013 at 4:29 PM, Andy W [via SPSSX Discussion]
<[hidden email]> wrote:

> I'm totally cool with a guest post, although my readership isn't anything to
> brag about! Just shoot me an email (not sure how the logistics work in
> wordpress but I'm sure I can figure it out).
>
> I was going to say something about Ryan's code on estimating Rasch models on
> my blog as well (sometime). That is a perfect example, instead of making a
> question linger on the list-serve better to get it out there on a website
> (and not have good things lost to the wind if Nabble goes away tomorrow).
> Alot of people have code posted on the web to the same effect (e.g. Marta,
> Bruce, Raynald), but I enjoy blogging about it in addition to that.
>
> Offer for guest posting on my blog is up to anyone. In the end I would like
> a SPSS blog roll like there is for R-bloggers, so make your own blogs! (And
> let me know about it)
>
> Andy
>
> ps - Those who consult, a blog is an excellent way to bring notice to your
> work. It may be purely coincidental but I got a pretty good gig with some
> people from the public policy dept. at my University that wouldn't have
> otherwise known about me.
> Andy W
> [hidden email]
> http://andrewpwheeler.wordpress.com/
>
>
> ________________________________
> If you reply to this email, your message will be added to the discussion
> below:
> http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718673.html
> To unsubscribe from One tail test, click here.
> NAML

---

View this message in context: Re: One tail test
Sent from the SPSSX Discussion mailing list archive at 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

[David Marso]

OH CRAP!!
I thought I was talking to Andy *alone*).
Must check the email headers more carefully ;-)
--

Art Kendall wrote

IIRC there
          was a discussion list or newsgroup back in
            the mid 80's.  Was it on BITNET? It was
      something later than ARPAnet or DECnet.
     
      It might have been one of the early competitors to L-Soft's
      LISTSERV.
      Art Kendall
Social Research Consultants
      On 3/14/2013 5:35 PM, David Marso wrote:
   
    I am actually on a project that came word of mouth
      from another client
     
      who I found from an SPSS listing.  All I had to do was link to my
     
      profile and it was !JOB IS YOURS!.
     
      Actually, Nabble is only one archive (UGA is the original :
     
      http://listserv.uga.edu/cgi-bin/wa?A0=spssx-l&D=0&F=&H=0&O=T&S=&T=1 
     
      Unfortunately a ton of stuff somehow got nuked!
     
      It appears to begin in Feb 1996 but that is oh so wrong!
     
      Or maybe comp.soft-sys.stat.spss (which has been spammed to death)
     
      Upon inspection it appears that  is nuked back to 1996 as well.  I
     
      clearly recall posting numerous gems in the early 1990's and was
     
      nearly out of support in 1996.  Nowdays everything I post of value
     
      accumulates on my HD.
     
      Yeah, If it is of any value it can be recreated!
     
      Let's see how the time works in coming weeks.
     
      BTW: Are you on LinkedIn?
     
      David
     
     
      On Thu, Mar 14, 2013 at 4:29 PM, Andy W [via SPSSX Discussion]
     
      < [hidden email] >
      wrote:
     
        > I'm totally cool with a guest post, although my readership
        isn't anything to
       
        > brag about! Just shoot me an email (not sure how the
        logistics work in
       
        > wordpress but I'm sure I can figure it out).
       
        >
       
        > I was going to say something about Ryan's code on
        estimating Rasch models on
       
        > my blog as well (sometime). That is a perfect example,
        instead of making a
       
        > question linger on the list-serve better to get it out
        there on a website
       
        > (and not have good things lost to the wind if Nabble goes
        away tomorrow).
       
        > Alot of people have code posted on the web to the same
        effect (e.g. Marta,
       
        > Bruce, Raynald), but I enjoy blogging about it in addition
        to that.
       
        >
       
        > Offer for guest posting on my blog is up to anyone. In the
        end I would like
       
        > a SPSS blog roll like there is for R-bloggers, so make your
        own blogs! (And
       
        > let me know about it)
       
        >
       
        > Andy
       
        >
       
        > ps - Those who consult, a blog is an excellent way to bring
        notice to your
       
        > work. It may be purely coincidental but I got a pretty good
        gig with some
       
        > people from the public policy dept. at my University that
        wouldn't have
       
        > otherwise known about me.
       
        > Andy W
       
        > [hidden email]
       
        > http://andrewpwheeler.wordpress.com/ 
        >
       
        >
       
        > ________________________________
       
        > If you reply to this email, your message will be added to
        the discussion
       
        > below:
       
        > http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718673.html 
        > To unsubscribe from One tail test, click here.
       
        > NAML
       
     
       Please
        reply to the list and not to my personal email.
       
        Those desiring my consulting or training services please feel
        free to email me.
     
     
      View this message in context: Re:
        One tail test
      Sent from the SPSSX
        Discussion mailing list archive  at 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

[Ryan]

Andy,

 

Feel free to let me know if you plan on posting something about the code I offered up to estimate the standard dichotomous Rasch model. There is so much more that I could have explored. For example, there MAY BE ways in which one could examine item fit, person fit, dimensionality, differential item function, etc, in the context of Rasch modeling in SPSS via a standard or random effects logistic regression that I have yet to explore. With the advent of the GENLINMIXED procedure, I bet that it is possible to fit even more complex 1-PL IRT models (e.g., treating persons and/or items as random effects; fitting a model in which persons are nested in higher level units--aka nested random effects).

 

I wrote an article in applied psychological measurement on this topic using the SAS GLIMMIX procedure about a year ago. I believe some of the capabilities I wrote about using the GLIMMIX procedure in SAS with respect to Rasch modeling MIGHT also be possible using the GENLINMIXED procedure in SPSS, although I cannot say for certain until I examine the capabilities of the GENLINMIXED procedure. Last I checked, it was quite behind the GLIMMIX procedure in SPSS.

 

Best,

 

Ryan

On Thu, Mar 14, 2013 at 4:29 PM, Andy W <[hidden email]> wrote:

I'm totally cool with a guest post, although my readership isn't anything to
brag about! Just shoot me an email (not sure how the logistics work in
wordpress but I'm sure I can figure it out).

I was going to say something about Ryan's code on estimating Rasch models on
my blog as well (sometime). That is a perfect example, instead of making a
question linger on the list-serve better to get it out there on a website
(and not have good things lost to the wind if Nabble goes away tomorrow).
Alot of people have code posted on the web to the same effect (e.g. Marta,
Bruce, Raynald), but I enjoy blogging about it in addition to that.

Offer for guest posting on my blog is up to anyone. In the end I would like
a SPSS blog roll like there is for R-bloggers, so make your own blogs! (And
let me know about it)

Andy

ps - Those who consult, a blog is an excellent way to bring notice to your
work. It may be purely coincidental but I got a pretty good gig with some
people from the public policy dept. at my University that wouldn't have
otherwise known about me.

-----
Andy W
[hidden email]
http://andrewpwheeler.wordpress.com/
--

View this message in context: http://spssx-discussion.1045642.n5.nabble.com/One-tail-test-tp5718623p5718673.html

Sent from the SPSSX Discussion mailing list archive at 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