Re: SPSSX-L Digest - 10 Feb 2008 to 11 Feb 2008 (#2008-42)

[Paul Ginns-2]

Hi all,

It's possible the reported effect size was a standardised mean
difference effect size appropriate for the paired t-test. While SPSS
doesn't report it, it can be calculated straightforwardly. See the
effect size calculator spreadsheet (dependent t-test option) at
http://mason.gmu.edu/~dwilsonb/ma.html (at the bottom of the page).
Other freeware for doing these calculations is described in the
following paper:

Algina, J., Keselman, H.J., & Penfield, R. D. (2005). Effect Sizes and
Their Intervals: The Two-Groups Correlated Case. Educational and
Psychological Measurement, 65, 241-258.

Cheers,

Paul

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

Date:    Mon, 11 Feb 2008 01:23:25 -0500
From:    Richard Ristow <[hidden email]>
Subject: Re: [Effect size from paired t-test]

At 10:25 PM 2/8/2008, John run wrote:

>I was reading a research article that reported the results of a
>paired-samples t-test. The author reported the significance (t(125)
>= 64.24, p<.0001) and then reported the effect size of.71. Is the
>correlation coefficient reported in the output generated by SPSS the
>effect size or was the effect size a separate calculation?

The effect size is almost certainly the estimated mean difference of
the two variables. In the following output (draft output) from SPSS's
T-TEST, it's the Mean under 'Paired Differences' in the Paired
Samples Test. (The code to generate this output is at the end of this
posting.)


Paired Samples Statistics
|----|--|------|---|--------------|---------------|
|    |  |Mean  |N  |Std. Deviation|Std. Error Mean|
|----|--|------|---|--------------|---------------|
|Pair|X1|5.9225|125|4.51181       |.40355         |
|1   |--|------|---|--------------|---------------|
|    |X2|6.6787|125|4.60598       |.41197         |
|----|--|------|---|--------------|---------------|

Paired Samples Correlations
|------|-------|---|-----------|----|
|      |       |N  |Correlation|Sig.|
|------|-------|---|-----------|----|
|Pair 1|X1 & X2|125|.976       |.000|
|------|-------|---|-----------|----|

Paired Samples Test
|------|-----|-------------------------------------|------|---|-------|
|      |     |Paired Differences                   |t     |df |Sig.   |
|      |     |------|-------|-------|--------------|      |   |(two-  |
|      |     |Mean  |Std.   |Std.   |95% Confidence|      |   |tailed)|
|      |     |      |Dev.   |Error  |Interval of   |      |   |       |
|      |     |      |       |of Mean|the Difference|      |   |       |
|      |     |      |       |       |------|-------|------|---|-------|
|      |     |      |       |       |Lower |Upper  |      |   |       |
|------|-----|------|-------|-------|------|-------|------|---|-------|
|Pair 1|X1-X2|-.7561|1.00418|.08982 |-.9339|-.5783 |-8.419|124|.000   |
|------|-----|------|-------|-------|------|-------|------|---|-------|

===================================
APPENDIX: Test data and T-test code
===================================
SET RNG = MT       /* 'Mersenne twister' random number generator  */ .
SET MTINDEX = 6820 /*  Providence, RI telephone book              */ .


NEW FILE.
INPUT PROGRAM.
.  NUMERIC CaseID (N3).
.  LEAVE   CaseID.
.  NUMERIC X1 X2  (F5.2).
.  LOOP    CaseID = 1 TO 125.
.     COMPUTE #Common = RV.NORMAL(0.1*CaseID,3).
.     COMPUTE #Delta  = RV.NORMAL(0.71,1).
.     COMPUTE X1      = #Common - #Delta/2.
.     COMPUTE X2      = #Common + #Delta/2.
.     END CASE.
.  END LOOP.
.  END FILE.
END INPUT PROGRAM.

T-TEST
   PAIRS = X1  WITH X2 (PAIRED)
   /CRITERIA = CI(.95)
   /MISSING = ANALYSIS.

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

[Manmit Shrimali-2]

This may sound like very simple and bad question but I am struggling to get answer on google. In discriminant analysis I realize that the function will always be total number of categories-1. Let say the dependent variable has four categories: north east west south. After running discriminant analysis I get only three functions. Where and how I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

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

[Zetu, Dan]

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

[Manmit Shrimali-2]

Thank you Dan for giving direction. I am using discriminant to find out if independent variables differ for e.g. marital status, gender, income by dependent variable (north, east, west, south). I thought discriminant analysis will give output like where the categories of independent variable fall to dependent variable for e.g. most of the male are in north while most of married are in east zone. Can I get such output in discriminant score? Sorry for sounding so silly.


-----Original Message-----
From:   SPSSX(r) Discussion on behalf of Zetu, Dan
Sent:   Tue 2/12/2008 9:16 PM
To:     [hidden email]
Cc:
Subject:             Re: Interpreting constant function

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

[Zetu, Dan]

Well, there are a few ways to ascertain that. First, look at
classification function coefficient magnitudes for each variable and the
higher the difference in variable coefficients for each category the
more significant that variable is in discriminating.

Or, outside discriminant analysis, cross-tab your significant variables
by group and you get the answer you need that way.

There may be other (presumably better) ways that I don't know of.

HTH.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 11:48 AM
To: [hidden email]
Subject: Re: Interpreting constant function

Thank you Dan for giving direction. I am using discriminant to find out
if independent variables differ for e.g. marital status, gender, income
by dependent variable (north, east, west, south). I thought discriminant
analysis will give output like where the categories of independent
variable fall to dependent variable for e.g. most of the male are in
north while most of married are in east zone. Can I get such output in
discriminant score? Sorry for sounding so silly.


-----Original Message-----
From:   SPSSX(r) Discussion on behalf of Zetu, Dan
Sent:   Tue 2/12/2008 9:16 PM
To:     [hidden email]
Cc:
Subject:             Re: Interpreting constant function

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

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

[Swank, Paul R]

Slight correction. The number of discriminant functions is equal to the
lesser of the number of groups less 1 and the number of dependent
variables. So if you have four categories you are discriminating and two
DVs, there would only be two discriminant functions.

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Zetu, Dan
Sent: Tuesday, February 12, 2008 9:46 AM
To: [hidden email]
Subject: Re: Interpreting constant function

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

[Zetu, Dan]

Thanks, Paul. I am not following though. Isn't the variable encoding the
groups the DV?

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Swank, Paul R
Sent: Wednesday, February 13, 2008 1:23 PM
To: [hidden email]
Subject: Re: Interpreting constant function

Slight correction. The number of discriminant functions is equal to the
lesser of the number of groups less 1 and the number of dependent
variables. So if you have four categories you are discriminating and two
DVs, there would only be two discriminant functions.

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Zetu, Dan
Sent: Tuesday, February 12, 2008 9:46 AM
To: [hidden email]
Subject: Re: Interpreting constant function

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

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

[Swank, Paul R]

Depends on how you look at it. Discriminant analysis is really just a
facet of MANOVA.

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston


-----Original Message-----
From: Zetu, Dan [mailto:[hidden email]]
Sent: Wednesday, February 13, 2008 1:43 PM
To: Swank, Paul R; [hidden email]
Subject: RE: Re: Interpreting constant function

Thanks, Paul. I am not following though. Isn't the variable encoding the
groups the DV?

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Swank, Paul R
Sent: Wednesday, February 13, 2008 1:23 PM
To: [hidden email]
Subject: Re: Interpreting constant function

Slight correction. The number of discriminant functions is equal to the
lesser of the number of groups less 1 and the number of dependent
variables. So if you have four categories you are discriminating and two
DVs, there would only be two discriminant functions.

Paul R. Swank, Ph.D.
Professor and Director of Research
Children's Learning Institute
University of Texas Health Science Center - Houston


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Zetu, Dan
Sent: Tuesday, February 12, 2008 9:46 AM
To: [hidden email]
Subject: Re: Interpreting constant function

Manmit:

The discriminant function that you see in the SPSS output is not the
function you need for classification purposes. The number of
discriminant functions is always number of categories - 1. There is
another set of functions, called Classification Functions (Fisher
functions), whose number is equal to the number of categories. If you
use those, you will get a score for each category and the record is
assigned to the category with the highest score. To obtain Fisher
coefficients, check Fisher's in the statistics tab of the discriminant
analysis.

Hope this helps.

-------------------------------
Dan Zetu
Analytical Consultant
R. L. Polk & Co.
248-728-7278
[hidden email]

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Manmit Shrimali
Sent: Tuesday, February 12, 2008 10:37 AM
To: [hidden email]
Subject: Interpreting constant function

This may sound like very simple and bad question but I am struggling to
get answer on google. In discriminant analysis I realize that the
function will always be total number of categories-1. Let say the
dependent variable has four categories: north east west south. After
running discriminant analysis I get only three functions. Where and how
I would interpret the missing group i.e. the 4th group?

Please advice.

Ms

=====================
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
*****************************************************************
This message has originated from R. L. Polk & Co.,
26955 Northwestern Highway, Southfield, MI 48033.
R. L. Polk & Co. sends various types of email
communications.  If this email message concerns the
potential licensing of a Polk product or service, and
you do not wish to receive further emails regarding Polk
products, forward this email to [hidden email]
with the word "remove" in the subject line.

The email and any files transmitted with it are confidential
and intended solely for the individual or entity to whom they
are addressed.

If you have received this email in error, please delete this
message and notify the Polk System Administrator at
[hidden email].
*****************************************************************

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

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