SPSSX Discussion

syntax to create T-scores from z-scores?

Classic

List

Threaded

7 messages Options

edunbar

syntax to create T-scores from z-scores?

Dear List members,

Okay I hate to ask, but having spent too much time trying to figue
this out on my laptop, what is the syntax to create T-scores from
z-scores?

Much thanks,

ed dunbar

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

Art Kendall

Re: syntax to create T-scores from z-scores?

open a clean instance of SPSS.
cut, paste, and run the following syntax.

data list list/id (f3), zscore(f4.2).
begin data.
001 -3.00
002 -2.58
003 -1.96
004 -1.68
005 0.00
006 3.00
007 2.58
008 1.96
009 1.68
010 -2.00
011 -1.00
012 1.00
013 2.00
end data.
sort cases by zscore.
numeric Tscore (f2).
compute Tscore = zscore +50 + (10*zscore).
list.

Art Kendall
Social Research Consultants
Celebrating the 60th Anniversary of the UN's Universal Declaration of
Human Rights

[hidden email] wrote:

> Dear List members,
>
> Okay I hate to ask, but having spent too much time trying to figue
> this out on my laptop, what is the syntax to create T-scores from
> z-scores?
>
> Much thanks,
>
> ed dunbar
>
> =====================
> 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
>
>

Art Kendall
Social Research Consultants

Marta Garcia-Granero

Re: syntax to create T-scores from z-scores?

Hi Art

I might be wrong, but I think that the conversion from Z scores to T
scores is:

Tscore=Zscore*10+50

Best regards,
Marta

> open a clean instance of SPSS.
> cut, paste, and run the following syntax.
>
> data list list/id (f3), zscore(f4.2).
> begin data.
> 001 -3.00
> 002 -2.58
> 003 -1.96
> 004 -1.68
> 005 0.00
> 006 3.00
> 007 2.58
> 008 1.96
> 009 1.68
> 010 -2.00
> 011 -1.00
> 012 1.00
> 013 2.00
> end data.
> sort cases by zscore.
> numeric Tscore (f2).
> compute Tscore = zscore +50 + (10*zscore).
> list.
>
> Art Kendall
> Social Research Consultants
> Celebrating the 60th Anniversary of the UN's Universal Declaration of
> Human Rights
>
>
> [hidden email] wrote:
>> Dear List members,
>>
>> Okay I hate to ask, but having spent too much time trying to figue
>> this out on my laptop, what is the syntax to create T-scores from
>> z-scores?
>>
>> Much thanks,
>>
>> ed dunbar
>>
>> =====================
>> 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
>

Jodene Fine

Re: syntax to create T-scores from z-scores?

Marta is correct. The conversion from a t-score to a z-score is t=z*10 + 50.
Conversions involving z-scores are special because the mean of a z-score is
0 and the SD is 1. The general formula for score conversion is:

Snew = ((Xold - Mold)/SDold)*SDnew+Mnew

Where:
Snew= new score
Xold = old score
Mold = old mean
SDold = old standard deviation
SDnew = new standard deviation
Mnew = new mean

Good luck,
Jodene

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Marta García-Granero
Sent: Wednesday, May 28, 2008 5:19 AM
To: [hidden email]
Subject: Re: syntax to create T-scores from z-scores?

Hi Art

I might be wrong, but I think that the conversion from Z scores to T
scores is:

Tscore=Zscore*10+50

Best regards,
Marta

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

Art Kendall

Re: syntax to create T-scores from z-scores?

In reply to this post by Marta Garcia-Granero

You are right.

Art

Marta García-Granero wrote:

> Hi Art
>
> I might be wrong, but I think that the conversion from Z scores to T
> scores is:
>
> Tscore=Zscore*10+50
>
> Best regards,
> Marta
>
>> open a clean instance of SPSS.
>> cut, paste, and run the following syntax.
>>
>> data list list/id (f3), zscore(f4.2).
>> begin data.
>> 001 -3.00
>> 002 -2.58
>> 003 -1.96
>> 004 -1.68
>> 005 0.00
>> 006 3.00
>> 007 2.58
>> 008 1.96
>> 009 1.68
>> 010 -2.00
>> 011 -1.00
>> 012 1.00
>> 013 2.00
>> end data.
>> sort cases by zscore.
>> numeric Tscore (f2).
>> compute Tscore = zscore +50 + (10*zscore).
>> list.
>>
>> Art Kendall
>> Social Research Consultants
>> Celebrating the 60th Anniversary of the UN's Universal Declaration of
>> Human Rights
>>
>>
>> [hidden email] wrote:
>>> Dear List members,
>>>
>>> Okay I hate to ask, but having spent too much time trying to figue
>>> this out on my laptop, what is the syntax to create T-scores from
>>> z-scores?
>>>
>>> Much thanks,
>>>
>>> ed dunbar
>>>
>>> =====================
>>> 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
>
>

Art Kendall
Social Research Consultants

Adam Troy

Case Matching Syntax

In reply to this post by Jodene Fine

Hi all,

I would appreciate anyone's assistance with this. I have a dataset of 300
students who responded to a 10-item true/false inventory. Each row is a
student, and each column (10 columns) is a student's response to a
particular inventory item (0 = true, 1 = false).

What I need to do is divide the sample into optimal groupings of either 3 or
4 students, that are most similar to each other based on their responses to
the questionnaire. I want to maximize overall sample "match" such that (for
example) within each group everyone would match on 5 items versus an
approach where some groups match on all 10 items and others match on 0
items.

Caveats: (1) some of the inventory items are more important to match on than
others, is there any method of matching first on some than others?
(2) There are two inventory items where "not matching" is preferred, how can
this be integrated into the syntax.

Any guidance to accomplish this through SPSS (perhaps syntax) is greatly
appreciated. Raw data can be provided on request.

Thanks,

Adam

______________________
Adam B. Troy, Ph.D.
Director of Research
GoalQuest
Enrollment and Retention Services Divisions

EducationDynamics
Chart Your Course

[hidden email]
Tel 201.377.3337
Fax 201.653.2971

www.educationdynamics.com

Please note our new mailing address:
111 River Street, Suite 1201
Hoboken, NJ 07030

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

Re: Case Matching Syntax

At 11:34 AM 5/28/2008, Adam B. Troy wrote:

>I have a dataset of 300 students who responded to a 10-item
>true/false inventory. Each row is a student, and each column (10
>columns) is a student's response to a particular inventory item (0 =
>true, 1 = false).
>
>I need to divide the sample into optimal groupings of either 3 or 4
>students, that are most similar to each other based on their responses

Cluster 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