Re: SPSSX-L Digest - 12 Sep 2006 to 13 Sep 2006 (#2006-253): computing chi-square from existing table data

classic Classic list List threaded Threaded
1 message Options
Reply | Threaded
Open this post in threaded view
|

Re: SPSSX-L Digest - 12 Sep 2006 to 13 Sep 2006 (#2006-253): computing chi-square from existing table data

Leslie Horst
Thank you Dominic.  I had gotten hung up in thinking you had to weight some variable by something, rather than just being able to say "weight by [whatever]"

For anyone else's information, here's the syntax that worked:

[after getting the file]

weight  by relatecount.
crosstabs     tables = freqrelate by presyears  /cells = count row col / statistics = chisq.
weight by meetcount.
crosstabs tables = freqmeet by presyears/cells = count row col / statistics = chisq.

Both tables exactly agreed with what I'd gotten by going back to the original data.


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Automatic digest processor
Sent: Thursday, September 14, 2006 12:05 AM
To: Recipients of SPSSX-L digests
Subject: SPSSX-L Digest - 12 Sep 2006 to 13 Sep 2006 (#2006-253)

There are 13 messages totalling 914 lines in this issue.

Topics of the day:

  1. Interesting links (in my opinion) (3)
  2. SAS formats ??? (3)
  3. calculations between rows solved
  4. SAS To SPSS Syntax help
  5. Chi-square
  6. evaluation multicollineary in Cox regression
  7. reading korean fonts in spss
  8. computing chi-square from existing table data (2)

----------------------------------------------------------------------

Date:    Wed, 13 Sep 2006 09:06:11 +0200
From:    =?ISO-8859-15?B?TWFydGEgR2FyY+1hLUdyYW5lcm8=?= <[hidden email]>
Subject: Re: Interesting links (in my opinion)

Hi Edward

EB> I am interested in buying the following books "Statistics at Square One"
EB> and "Research Methods II: Multivariate Analysis" .

These are on-line books... As a matter of fact, I don't think the
second one exists in hard copy (unless you take the time to download
every pdf file and print them, as I did).
http://www.oxfordjournals.org/tropej/online/ma.html

The first one, Statistics at Square One, can be viewed on line too
http://bmj.bmjjournals.com/collections/statsbk/index.shtml
Until I bought it, I made the effort of printing to PDF every chapter
and putting all chapter together in one single file. It's quite easy
(if you have Acrobat PDF Writer installed in you computer).

A printed version (10th edition, instead of 9th edition, the one on line)
can be bought at BMJ e-store:
http://www.bmjbookshop.com/shop/product_display.asp?&SiteLanguage=ENG&productid=0727915525

Once there, if you are really insterested in buying books, then take a
look also at "Statistics at Square Two" (I bought it too, quite good)


Regards,
Marta

EB> -----Original Message-----
EB> From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
EB> Marta García-Granero
EB> Sent: Thursday, June 29, 2006 12:17 PM
EB> To: [hidden email]
EB> Subject: Interesting links (in my opinion)

EB> Hi everybody:

EB> Since a lot of questions to this list concern theoretical issues, I
EB> thought that these links could be interesting for a lot of people. The
EB> first one reviews basic stats, while the second focuses on more
EB> advanced methods.

EB> Statistics at Square One
EB> =========================
EB> 1 Data display and summary
EB> 2 Mean and standard deviation
EB> 3 Populations and samples
EB> 4 Statements of probability and confidence intervals
EB> 5 Differences between means: type I and type II errors and power
EB> 6 Differences between percentages and paired alternatives
EB> 7 The t tests
EB> 8 The chi-squared tests
EB> 9 Exact probabilty test
EB> 10 Rank score tests
EB> 11 Correlation and regression
EB> 12 Survival analysis
EB> 13 Study design and choosing a statistical test

EB> Available as on-line book (HTML) at:
EB> http://bmj.bmjjournals.com/collections/statsbk/index.shtml

EB> SPSS syntax code (written by me and kindly hosted by King Douglas) for
EB> the book exercises and examples is available at:
EB> http://www.kingdouglas.com/SPSS/DiverseCultures/Marta/Code/BMJ%20-%20Stats%20Square%20One.txt
EB> (if the link is broken in two by my mail program, copy and paste both
EB> pieces together...)

EB> ------------------------------------------------------------------------
EB> Research Methods II: Multivariate Analysis
EB> ============================================
EB> Chapter 1: The Scientific Method
EB> Chapter 2: Simple Linear Regression
EB> Chapter 3: Multiple Regression Analysis
EB> Chapter 4: Multiple Regression in Practice
EB> Chapter 5: Regression Diagnostics
EB> Chapter 6: Analysis of Cross-Sectional Studies
EB> Chapter 7: One-way Analysis of Variance
EB> Chapter 8: Two-way Analysis of Variance
EB> Chapter 9: Factorial Designs
EB> Chapter 10: Repeat-measures Designs
EB> Chapter 11: Logistic Regression
EB> Chapter 12: Survival Analysis
EB> Chapter 13: Poisson Regression Analysis
EB> Chapter 14: Analysing Categorical Data: Log-linear analysis

EB> Available as on-line book (PDF) at:

EB> http://www.oxfordjournals.org/tropej/online/ma.html
EB> (they also have a collection of PowerPoint slides for each chapter)

EB> Although I haven't written SPSS code for the second book (I just
EB> discovered it today!), give me some time and (after my holidays), I'll
EB> have it available.


EB> I hope you find them useful.

------------------------------

Date:    Wed, 13 Sep 2006 09:44:05 +0200
From:    Mark Webb <[hidden email]>
Subject: Re: Interesting links (in my opinion)

Is the second book available on-line like the first ?
Or only in print format ?

Regards
Mark

----- Original Message -----
From: "Marta García-Granero" <[hidden email]>
To: <[hidden email]>
Sent: Wednesday, September 13, 2006 9:06 AM
Subject: Re: Interesting links (in my opinion)


> Hi Edward
>
> EB> I am interested in buying the following books "Statistics at Square
> One"
> EB> and "Research Methods II: Multivariate Analysis" .
>
> These are on-line books... As a matter of fact, I don't think the
> second one exists in hard copy (unless you take the time to download
> every pdf file and print them, as I did).
> http://www.oxfordjournals.org/tropej/online/ma.html
>
> The first one, Statistics at Square One, can be viewed on line too
> http://bmj.bmjjournals.com/collections/statsbk/index.shtml
> Until I bought it, I made the effort of printing to PDF every chapter
> and putting all chapter together in one single file. It's quite easy
> (if you have Acrobat PDF Writer installed in you computer).
>
> A printed version (10th edition, instead of 9th edition, the one on line)
> can be bought at BMJ e-store:
> http://www.bmjbookshop.com/shop/product_display.asp?&SiteLanguage=ENG&productid=0727915525
>
> Once there, if you are really insterested in buying books, then take a
> look also at "Statistics at Square Two" (I bought it too, quite good)
>
>
> Regards,
> Marta
>
> EB> -----Original Message-----
> EB> From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf
> Of
> EB> Marta García-Granero
> EB> Sent: Thursday, June 29, 2006 12:17 PM
> EB> To: [hidden email]
> EB> Subject: Interesting links (in my opinion)
>
> EB> Hi everybody:
>
> EB> Since a lot of questions to this list concern theoretical issues, I
> EB> thought that these links could be interesting for a lot of people. The
> EB> first one reviews basic stats, while the second focuses on more
> EB> advanced methods.
>
> EB> Statistics at Square One
> EB> =========================
> EB> 1 Data display and summary
> EB> 2 Mean and standard deviation
> EB> 3 Populations and samples
> EB> 4 Statements of probability and confidence intervals
> EB> 5 Differences between means: type I and type II errors and power
> EB> 6 Differences between percentages and paired alternatives
> EB> 7 The t tests
> EB> 8 The chi-squared tests
> EB> 9 Exact probabilty test
> EB> 10 Rank score tests
> EB> 11 Correlation and regression
> EB> 12 Survival analysis
> EB> 13 Study design and choosing a statistical test
>
> EB> Available as on-line book (HTML) at:
> EB> http://bmj.bmjjournals.com/collections/statsbk/index.shtml
>
> EB> SPSS syntax code (written by me and kindly hosted by King Douglas) for
> EB> the book exercises and examples is available at:
> EB>
> http://www.kingdouglas.com/SPSS/DiverseCultures/Marta/Code/BMJ%20-%20Stats%20Square%20One.txt
> EB> (if the link is broken in two by my mail program, copy and paste both
> EB> pieces together...)
>
> EB> ------------------------------------------------------------------------
> EB> Research Methods II: Multivariate Analysis
> EB> ============================================
> EB> Chapter 1: The Scientific Method
> EB> Chapter 2: Simple Linear Regression
> EB> Chapter 3: Multiple Regression Analysis
> EB> Chapter 4: Multiple Regression in Practice
> EB> Chapter 5: Regression Diagnostics
> EB> Chapter 6: Analysis of Cross-Sectional Studies
> EB> Chapter 7: One-way Analysis of Variance
> EB> Chapter 8: Two-way Analysis of Variance
> EB> Chapter 9: Factorial Designs
> EB> Chapter 10: Repeat-measures Designs
> EB> Chapter 11: Logistic Regression
> EB> Chapter 12: Survival Analysis
> EB> Chapter 13: Poisson Regression Analysis
> EB> Chapter 14: Analysing Categorical Data: Log-linear analysis
>
> EB> Available as on-line book (PDF) at:
>
> EB> http://www.oxfordjournals.org/tropej/online/ma.html
> EB> (they also have a collection of PowerPoint slides for each chapter)
>
> EB> Although I haven't written SPSS code for the second book (I just
> EB> discovered it today!), give me some time and (after my holidays), I'll
> EB> have it available.
>
>
> EB> I hope you find them useful.
>
> __________ NOD32 1.1753 (20060912) Information __________
>
> This message was checked by NOD32 antivirus system.
> http://www.eset.com
>
>

------------------------------

Date:    Wed, 13 Sep 2006 11:58:18 +0300
From:    Samuel Solomon <[hidden email]>
Subject: SAS formats ???

Hi list,

  I hope you would allow me to ask SAS related queries. Here is a SAS
code..=20

=20

Proc format;

picture mf

     0             =3D'99999999999999'

     0 < -high     =3D'99999999999999';

picture nf

     0            =3D '        '

     0 < - high   =3D '009,99' (mult=3D100);

=20

Does it ring any bells to you? What does it me? I wish to convert it to
a meaningful SPSS syntax.

=20

Thanks,

Samuel.

=20

------------------------------

Date:    Wed, 13 Sep 2006 13:50:22 +0100
From:    Paul Mcgeoghan <[hidden email]>
Subject: Re: calculations between rows solved

Edward,

This is the syntax I used:

AGGREGATE
  /OUTFILE='D:/aggr.sav'
  /BREAK=id
  /date_first = FIRST(date) /date_last = LAST(date).

GET
  FILE='D:\aggr.sav'.
DATASET NAME DataSet2 WINDOW=FRONT.

COMPUTE days = DATEDIFF(date_last,date_first,"days") .
EXECUTE .

Paul

==================
Paul McGeoghan,
Application support specialist (Statistics and Databases),
University Infrastructure Group (UIG),
Information Services,
Cardiff University.
Tel. 02920 (875035).

==================
Paul McGeoghan,
Application support specialist (Statistics and Databases),
University Infrastructure Group (UIG),
Information Services,
Cardiff University.
Tel. 02920 (875035).

>>> "Edward Boadi" <[hidden email]> 12/09/2006 14:10 >>>
Paul, I am interested in knowing (syntax) how that
was done.
Regards.
Edward.

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Paul Mcgeoghan
Sent: Tuesday, September 12, 2006 6:03 AM
To: [hidden email]
Subject: calculations between rows solved


Figured it out using Data Aggregate,
I can specify the 1st row and the last row and create a new aggregated dataset
from that where I can then compute the difference between the 2 dates.

Hi,

I have a customer with data as follows:
      ID        Date
         1      11-DEC-2001
       1        21-DEC-2001
       1        30-DEC-2001
       2        10-JAN-2000
       2        12-JAN-2000
       2        15-FEB-2000
       2        20-MAR-2000
       2        26-APR-2000

He wants to have:
ID  Days
1   19
2   107

What is best way to do this?
Thanks,
Paul

==================
Paul McGeoghan,
Application support specialist (Statistics and Databases),
University Infrastructure Group (UIG),
Information Services,
Cardiff University.
Tel. 02920 (875035).

------------------------------

Date:    Wed, 13 Sep 2006 15:10:38 +0200
From:    Spousta Jan <[hidden email]>
Subject: Re: SAS formats ???

Hi Samuel,

I am not an expert in SAS formats and it is hard to reproduce them
correctly in SPSS. I think that the first format is simply

FORMATS variable list (F14).

in the SPSS language. (Of course, in SPSS you cannot create formats in
advance, without applying them on variables.)

The second case is more difficult - it is something like

COMPUTE x = x * 100 /* the MULT option in SAS */ .
FORMATS x (COMMA6.2).

Moreover SAS replaces zero with blanks - hard to code in SPSS without
transforming the numbers into strings.

Hope this helps somehow,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Samuel Solomon
Sent: Wednesday, September 13, 2006 10:58 AM
To: [hidden email]
Subject: SAS formats ???

Hi list,

  I hope you would allow me to ask SAS related queries. Here is a SAS
code..



Proc format;

picture mf

     0             ='99999999999999'

     0 < -high     ='99999999999999';

picture nf

     0            = '        '

     0 < - high   = '009,99' (mult=100);



Does it ring any bells to you? What does it me? I wish to convert it to
a meaningful SPSS syntax.



Thanks,

Samuel.

------------------------------

Date:    Wed, 13 Sep 2006 09:31:40 -0400
From:    Edward Boadi <[hidden email]>
Subject: Re: Interesting links (in my opinion)

Thank you , Marta & Angshu .

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Marta García-Granero
Sent: Wednesday, September 13, 2006 3:06 AM
To: [hidden email]
Subject: Re: Interesting links (in my opinion)


Hi Edward

EB> I am interested in buying the following books "Statistics at Square One"
EB> and "Research Methods II: Multivariate Analysis" .

These are on-line books... As a matter of fact, I don't think the
second one exists in hard copy (unless you take the time to download
every pdf file and print them, as I did).
http://www.oxfordjournals.org/tropej/online/ma.html

The first one, Statistics at Square One, can be viewed on line too
http://bmj.bmjjournals.com/collections/statsbk/index.shtml
Until I bought it, I made the effort of printing to PDF every chapter
and putting all chapter together in one single file. It's quite easy
(if you have Acrobat PDF Writer installed in you computer).

A printed version (10th edition, instead of 9th edition, the one on line)
can be bought at BMJ e-store:
http://www.bmjbookshop.com/shop/product_display.asp?&SiteLanguage=ENG&productid=0727915525

Once there, if you are really insterested in buying books, then take a
look also at "Statistics at Square Two" (I bought it too, quite good)


Regards,
Marta

EB> -----Original Message-----
EB> From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
EB> Marta García-Granero
EB> Sent: Thursday, June 29, 2006 12:17 PM
EB> To: [hidden email]
EB> Subject: Interesting links (in my opinion)

EB> Hi everybody:

EB> Since a lot of questions to this list concern theoretical issues, I
EB> thought that these links could be interesting for a lot of people. The
EB> first one reviews basic stats, while the second focuses on more
EB> advanced methods.

EB> Statistics at Square One
EB> =========================
EB> 1 Data display and summary
EB> 2 Mean and standard deviation
EB> 3 Populations and samples
EB> 4 Statements of probability and confidence intervals
EB> 5 Differences between means: type I and type II errors and power
EB> 6 Differences between percentages and paired alternatives
EB> 7 The t tests
EB> 8 The chi-squared tests
EB> 9 Exact probabilty test
EB> 10 Rank score tests
EB> 11 Correlation and regression
EB> 12 Survival analysis
EB> 13 Study design and choosing a statistical test

EB> Available as on-line book (HTML) at:
EB> http://bmj.bmjjournals.com/collections/statsbk/index.shtml

EB> SPSS syntax code (written by me and kindly hosted by King Douglas) for
EB> the book exercises and examples is available at:
EB> http://www.kingdouglas.com/SPSS/DiverseCultures/Marta/Code/BMJ%20-%20Stats%20Square%20One.txt
EB> (if the link is broken in two by my mail program, copy and paste both
EB> pieces together...)

EB> ------------------------------------------------------------------------
EB> Research Methods II: Multivariate Analysis
EB> ============================================
EB> Chapter 1: The Scientific Method
EB> Chapter 2: Simple Linear Regression
EB> Chapter 3: Multiple Regression Analysis
EB> Chapter 4: Multiple Regression in Practice
EB> Chapter 5: Regression Diagnostics
EB> Chapter 6: Analysis of Cross-Sectional Studies
EB> Chapter 7: One-way Analysis of Variance
EB> Chapter 8: Two-way Analysis of Variance
EB> Chapter 9: Factorial Designs
EB> Chapter 10: Repeat-measures Designs
EB> Chapter 11: Logistic Regression
EB> Chapter 12: Survival Analysis
EB> Chapter 13: Poisson Regression Analysis
EB> Chapter 14: Analysing Categorical Data: Log-linear analysis

EB> Available as on-line book (PDF) at:

EB> http://www.oxfordjournals.org/tropej/online/ma.html
EB> (they also have a collection of PowerPoint slides for each chapter)

EB> Although I haven't written SPSS code for the second book (I just
EB> discovered it today!), give me some time and (after my holidays), I'll
EB> have it available.


EB> I hope you find them useful.

------------------------------

Date:    Wed, 13 Sep 2006 09:24:18 -0400
From:    Richard Ristow <[hidden email]>
Subject: Re: SAS To SPSS Syntax help

At 07:23 PM 9/11/2006, Eddy Crane wrote:

>I have a SAS file to run a frequency and reach analysis that I am
>trying to recreate into SPSS.
>
>Where I am having the most trouble is with the "IF THEN" statements.
>Mainly because of the {bskcnt} and {a} statements.  also, the SAS code
>goes on and on with the else if statements for &bsk=n and I figure
>there would be some way to loop it to make it more efficient.

The {bskcnt} and {a} are array references, with 'bskcnt' and 'a' as the
indices.

Your code seems to mark, in array or vector 'reach', all combinations
of &bsk elements from &fl inputs for which any of the elements has a
non-zero value.

I've translated the SAS to SPSS for &bsk=3. You could write it without
a separate block for each value of &bsk, but it means reworking the
logic. I'm not going to do that, at least not unless you say more of
what this is about. The result looks clumsy to use, as it stands.

I've changed the SAS code to SPSS comments, and followed them by the
equivalent SPSS. The SPSS code is not tested.

*  The code has been 'unwrapped' from a SAS macro, whose only working .
*  content seems to be the MACRO statement itself:                    .
*
*     %MACRO CHOC(fl,bsk);
*
*  In the SPSS code, macro parameters &fl and &bsk have been replaced .
*  by SPSS variables @f1 and @bsk.                                    .
*  THIS CODE WILL NOT BE TESTED, OR RUN IN ANY WAY                    .

*  Only the block for &bsk=3 is translated.                           .

*SAS  data b;    --- .
*SAS  infile indat lrecl=10 recfm=v missover;    --- .
*SAS  input @1 (input1-input10)  (1.);    --- .
.     DATA LIST FILE=indat FIXED
               /input1 TO input10 (10F1).

*SAS  run;    --- .
*===  Irrelevant; equivalent to SPSS's EXECUTE .

*SAS  data c;    --- .
*SAS  set b;     --- .
*===  Relevant only if you want to keep input dataset b unchanged for .
*===  future use. If you do,                                          .
.     DATASET COPY b   /* SPSS 14 and later   */.
*===  or                                                              .
.     SAVE OUTFILE=b.  /* SPSS 13 and earlier */

*SAS  array input{12} input1-input12;    --- .
.     NUMERIC input11 input12 (F1).
.     VECTOR input=input1 TO input10.
*---  Remark: You have input1 TO input10 in your input. I've     --- .
*     declared input11 and input 12.                             --- .
*     The VECTOR works only if input1-input12 are contiguous in  --- .
*     the file. If this declaration doesn't make them            --- .
*     contiguous, you can declare all 12 contiguous in an        --- .
*     INPUT PROGRAM, then readonly the first 10.                 --- .

*SAS  array  reach{1000} reach1-reach1000;    --- .
.     VECTOR reach(1000,F3).
*SAS  array  fqcy{1000}  fqcy1-fqcy1000;      --- .
.     VECTOR fqcy(1000,F3).

<code for &bsk=1 and &bsk=2 omitted>

*SAS  else if &bsk=3 then do;    --- .
.     ELSE IF @BSK EQ 3.

*SAS     bskcnt=0;              --- .
.        COMPUTE bskcnt = 0.
*SAS     do a=1 to &fl;         --- .
.           LOOP a=1 TO @f1.

*SASb       bcnt=a+1;             --- .
*SAS        do b=bcnt to &fl;     --- .
*===        Either                    .
.           COMPUTE  bcnt = a+1.
.           LOOP b = bcnt TO @f1.
*===        or                        .
*           LOOP b = a+1 TO @f1.

*SASb          ccnt=b+1;            --- .
*SAS           do c=ccnt to &fl;    --- .
*===           Either                   .
.              COMPUTE  ccnt=b+1
.              LOOP c = ccnt TO @f1.
*===           or                       .
.              LOOP c = b+1  TO @f1     .


*SAS              bskcnt=bskcnt+1;    --- .
.                 COMPUTE bskcnt = bskcnt+1.
*---              Remark: so 'bskcnt' counts total loop iterations  .

*SAS              fqcy{bskcnt}=input{a}+input{b}+input{c};      --- .
.                 COMPUTE fqcy(bskcnt)
                              = input(a)+input(b)+input(c).
*---              Remark: The a/b/c nested loop adds all        --- .
*---              possible triples of different values from     --- .
*---              the inputs, and stores each in its own array  --- .
*---              element ('vector element' in SPSS).           --- .

*SAS              if fqcy{bskcnt} ge 1 then reach{bskcnt}=1;    --- .
*SASb                                  else reach{bskcnt}=0;    --- .
*SAS              if fqcy{bskcnt}=0    then fqcy{bskcnt}=.;     --- .
*---              Question: Are the numbers in array 'fqcy'     --- .
*---              ever used again? If not, there's no reason    --- .
*---              it need by an array; a scalar would do fine.  --- .

*SAS           end /* loop c */;  --- .
.              END LOOP /* c */.
*SAS        end  /* loop b */;  --- .
.           END LOOP  /* b */.
*SAS     end   /* loop a */;  --- .
.        END LOOP   /* a *.

*SAS  end;                  ---  .
*===  Omit in SPSS, if an ELSE IF or ELSE follows.      --- .
*===  If it's the last clause, then                     --- .
.     END IF.

------------------------------

Date:    Wed, 13 Sep 2006 10:06:51 -0400
From:    Richard Ristow <[hidden email]>
Subject: Re: Chi-square

At 12:06 PM 9/12/2006, Seiph wrote:

>I have a very basic stats question about how to determine differences
>between groups using a the chi-square statistic available during a
>cross-tabulation analysis in spss.
>
>I ran a crosstabs and selected the chi-square statistic. Here is the
>pasted syntax:
>CROSSTABS
>   /TABLES= STATUS BY EDUCLV
>   /FORMAT= AVALUE TABLES
>   /STATISTIC=CHISQ
>   /CELLS= COUNT ROW COLUMN TOTAL
>   /COUNT ROUND CELL .
>
>The output tells me that there is a significant difference between
>groups. So here's my basic question: how do I find out exactly where
>that difference is?

Try adding, to subcommand CELLS, EXPECTED and ASRESID. (That puts a lot
of numbers in each cell; you may want to drop one or more of the
percents.)

EXPECTED is the expected number in the cell, if the null hypothesis is
correct, and ASRESID is the adjusted standardized difference from
observed and expected. Look for cells with ASRESID greater than 2 in
absolute value; that will show cells where the difference is most
important.

(Marta, do I have this wrong in any way?)

------------------------------

Date:    Wed, 13 Sep 2006 10:11:34 -0400
From:    Christian Bautista <[hidden email]>
Subject: evaluation multicollineary in Cox regression

Hi all,

Is there any way to evaluate multicollinearity or collinearity after
running Cox regression analysis with SPSS?

Thansk,

/Christian

------------------------------

Date:    Wed, 13 Sep 2006 10:05:12 -0500
From:    "Swank, Paul R" <[hidden email]>
Subject: Re: SAS formats ???

The picture statement in SAS works like in COBOL. It merely provides a
format for printing. nf = to 0 is printed as blanks while nf>0 means
multiply by 100 and print with two decimals.


Paul R. Swank, Ph.D.
Professor, Developmental Pediatrics
Director of Research, Children's Learning Institute
Medical School
University of Texas Health Science Center at Houston

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Samuel Solomon
Sent: Wednesday, September 13, 2006 3:58 AM
To: [hidden email]
Subject: SAS formats ???

Hi list,

  I hope you would allow me to ask SAS related queries. Here is a SAS
code..



Proc format;

picture mf

     0             ='99999999999999'

     0 < -high     ='99999999999999';

picture nf

     0            = '        '

     0 < - high   = '009,99' (mult=100);



Does it ring any bells to you? What does it me? I wish to convert it to
a meaningful SPSS syntax.



Thanks,

Samuel.

------------------------------

Date:    Wed, 13 Sep 2006 10:20:47 -0400
From:    Merlin Marshall <[hidden email]>
Subject: reading korean fonts in spss

Greetings,

I am working with a Korean longitudinal data survey.  The data come to me
with Korean labels, which shows up as garbage on my Win XP-Pro system.  I'd
like them to show up correctly in my SPSS version 14.0, so that I can put
them into something like Google Translate and make English labels.  I have
Korean fonts installed on my XP computer and Word documents display Korean
correctly, but SPSS does not.  Has anyone tried sucessfully to get Asian
fonts to display in English SPSS?

thanks,

Merlin Marshall
Data Wrangler
Center for Human Resource Research
Ohio State University
Columbus Ohio

------------------------------

Date:    Wed, 13 Sep 2006 18:03:16 -0400
From:    Leslie Horst <[hidden email]>
Subject: computing chi-square from existing table data

Suppose I have data of the following form:

freqrelate      presyears       relatecount
1               1               111
1               2               122
1               3               112
1               4               113
2               1               63
2               2               61
2               3               76
2               4               85

Freqrelate is a classification variable as is presyears - I took these
data from an existing table.  Relatecount is the cell count for each of
the cells in the 2x4 table defined by freqrelate and presyears.  I
remember once (long ago) having a similar situation and being somehow
able to use SPSS to weight by the variable with the cell frequencies in
it and get the table statistics, but I can no longer remember how.  I
took a quick look in Raynald's Tools, but didn't find anything.  How can
I do this in syntax?  Thanks in advance.

Leslie Horst, Ph.D.
Maguire Associates, Inc.
5 Concord Farms
555 Virginia Rd., Suite 201
Concord, MA 01742

------------------------------

Date:    Wed, 13 Sep 2006 15:06:29 -0700
From:    Dominic Lusinchi <[hidden email]>
Subject: Re: computing chi-square from existing table data

Leslie, use the WEIGHT function: i.e. WEIGHT BY relatecount before the
CROSSTAB procedure.

Dominic Lusinchi
Statistician
Far West Research
Statistical Consulting
San Francisco, California
415-664-3032
www.farwestresearch.com
-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Leslie Horst
Sent: Wednesday, September 13, 2006 3:03 PM
To: [hidden email]
Subject: computing chi-square from existing table data

Suppose I have data of the following form:

freqrelate      presyears       relatecount
1               1               111
1               2               122
1               3               112
1               4               113
2               1               63
2               2               61
2               3               76
2               4               85

Freqrelate is a classification variable as is presyears - I took these
data from an existing table.  Relatecount is the cell count for each of
the cells in the 2x4 table defined by freqrelate and presyears.  I
remember once (long ago) having a similar situation and being somehow
able to use SPSS to weight by the variable with the cell frequencies in
it and get the table statistics, but I can no longer remember how.  I
took a quick look in Raynald's Tools, but didn't find anything.  How can
I do this in syntax?  Thanks in advance.

Leslie Horst, Ph.D.
Maguire Associates, Inc.
5 Concord Farms
555 Virginia Rd., Suite 201
Concord, MA 01742

------------------------------

End of SPSSX-L Digest - 12 Sep 2006 to 13 Sep 2006 (#2006-253)
**************************************************************