SPSSX Discussion

Significant difference - Means

Classic

List

Threaded

20 messages Options

Samir Omerovic

Significant difference - Means

Hi to all,

My first question is what test should I use if I need to assess the significance of the mean of some
subgroup in comparison the mean of the total population. For example I have 1000 respondents that
answered a question on 7-point scale. Their mean/average is 3,5. I have 500 students among those
1000 respondents with mean/average 2,9. I want to know if these values are significantly different.
What test should I use in SPSS?

Thanks to all in advance.

Samir

GfK BH

Centar za istrazivanje trzista i javnog mnijenja d.o.o.

Samir Omerović

Researcher

71 000 Sarajevo, Skenderija 44

Bosna i Hercegovina

Tel: +387 33 550 300

Fax: +387 33 444 226

www.gfk.ba

www.gfk.com

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

This e-mail (and any attachment/s) contains confidential and/or privileged information. If you are
not the intended recipient (or have received this e-mail in error) please notify the sender
immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the
material in this e-mail is strictly forbidden.

Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste osoba naznacena kao primalac
molimo odmah upozoriti posiljaoca i unistiti primljenu e-poruku. Neautorizirano kopiranje,
objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.

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

Spousta Jan

Re: Significant difference - Means

Hi Samir,

It is usually performed as a test students x rest of the group (=non-students), that is Independent-samples T-test in Analyze -> Compare means..., use the indicator of "studentness" as the Grouping Variable.

HTH

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Samir Omerović
Sent: Tuesday, December 19, 2006 9:06 AM
To: [hidden email]
Subject: Significant difference - Means

Hi to all,

My first question is what test should I use if I need to assess the significance of the mean of some subgroup in comparison the mean of the total population. For example I have 1000 respondents that answered a question on 7-point scale. Their mean/average is 3,5. I have 500 students among those 1000 respondents with mean/average 2,9. I want to know if these values are significantly different.
What test should I use in SPSS?

Thanks to all in advance.

Samir

GfK BH

Centar za istrazivanje trzista i javnog mnijenja d.o.o.

Samir Omerović

Researcher

71 000 Sarajevo, Skenderija 44

Bosna i Hercegovina

Tel: +387 33 550 300

Fax: +387 33 444 226

www.gfk.ba

www.gfk.com

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

This e-mail (and any attachment/s) contains confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorized copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i unistiti primljenu e-poruku. Neautorizirano kopiranje, objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.

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

JOHN ANTONAKIS

Re: Significant difference - Means

In reply to this post by Samir Omerovic

If you wiysh to test the mean of a particular group against the mean of a
particular value then you can use the single-sample t-test (from the Menus
Analyze/Compare Means/One sample t-test), or use the syntax:

T-TEST
/TESTVAL = 3.5
/MISSING = ANALYSIS
/VARIABLES = VAR00001
/CRITERIA = CI(.95) .

HTH,
J.

At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:

>Hi to all,
>
>
>
>My first question is what test should I use if I need to assess the
>significance of the mean of some
>subgroup in comparison the mean of the total population. For example I
>have 1000 respondents that
>answered a question on 7-point scale. Their mean/average is 3,5. I have
>500 students among those
>1000 respondents with mean/average 2,9. I want to know if these values are
>significantly different.
>What test should I use in SPSS?
>
>
>
>Thanks to all in advance.
>
>
>
>Samir
>
>
>
>GfK BH
>
>Centar za istrazivanje trzista i javnog mnijenja d.o.o.
>
>
>
>Samir Omeroviæ
>
>Researcher
>
>
>
>71 000 Sarajevo, Skenderija 44
>
>Bosna i Hercegovina
>
>
>
>
>
>Tel: +387 33 550 300
>
>Fax: +387 33 444 226
>
>www.gfk.ba
>
>www.gfk.com
>
>
>
>
>
>
>
>-----------------------------------------
>
>This e-mail (and any attachment/s) contains confidential and/or privileged
>information. If you are
>not the intended recipient (or have received this e-mail in error) please
>notify the sender
>immediately and destroy this e-mail. Any unauthorized copying,
>disclosure or distribution of the
>material in this e-mail is strictly forbidden.
>
>Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste osoba
>naznacena kao primalac
>molimo odmah upozoriti posiljaoca i unistiti primljenu e-poruku.
>Neautorizirano kopiranje,
>objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
>
>---------------------------------------
>
>

___________________________________

Prof. John Antonakis
Faculty of Management and Economics
University of Lausanne
Internef #527
CH-1015 Lausanne-Dorigny
Switzerland

Tel: ++41 (0)21 692-3438
Fax: ++41 (0)21 692-3305

http://www.hec.unil.ch/jantonakis
___________________________________

Spousta Jan

Re: Significant difference - Means

In reply to this post by Samir Omerovic

I am very sorry, but this would be wrong in this case. Because the number 3.5 is not a constant known with (almost) no error, but a sample statistics based on the 1000 respondents. Therefore the Independent-samples T-test should be used as I've written in my previous posting.

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John Antonakis
Sent: Tuesday, December 19, 2006 9:25 AM
To: [hidden email]
Subject: Re: Significant difference - Means

If you wiysh to test the mean of a particular group against the mean of a particular value then you can use the single-sample t-test (from the Menus Analyze/Compare Means/One sample t-test), or use the syntax:

T-TEST
/TESTVAL = 3.5
/MISSING = ANALYSIS
/VARIABLES = VAR00001
/CRITERIA = CI(.95) .

HTH,
J.

At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:

>Hi to all,
>
>
>
>My first question is what test should I use if I need to assess the
>significance of the mean of some subgroup in comparison the mean of the
>total population. For example I have 1000 respondents that answered a
>question on 7-point scale. Their mean/average is 3,5. I have 500
>students among those 1000 respondents with mean/average 2,9. I want to
>know if these values are significantly different.
>What test should I use in SPSS?
>
>
>
>Thanks to all in advance.
>
>
>
>Samir
>
>
>
>GfK BH
>
>Centar za istrazivanje trzista i javnog mnijenja d.o.o.
>
>
>
>Samir Omeroviæ
>
>Researcher
>
>
>
>71 000 Sarajevo, Skenderija 44
>
>Bosna i Hercegovina
>
>
>
>
>
>Tel: +387 33 550 300
>
>Fax: +387 33 444 226
>
>www.gfk.ba
>
>www.gfk.com
>
>
>
>
>
>
>
>-----------------------------------------
>
>This e-mail (and any attachment/s) contains confidential and/or
>privileged information. If you are not the intended recipient (or have
>received this e-mail in error) please notify the sender immediately
>and destroy this e-mail. Any unauthorized copying, disclosure or
>distribution of the material in this e-mail is strictly forbidden.
>
>Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
>osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
>unistiti primljenu e-poruku.
>Neautorizirano kopiranje,
>objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
>
>---------------------------------------
>
>

JOHN ANTONAKIS

Re: Significant difference - Means

In reply to this post by Samir Omerovic

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes
that 3.5 represents the population mean (with a sample of 1,000 the
standard error should be very small) then that is precisely the case in
which one would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that
the independent sample t-test is more appropriate (because it takes into
account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

>I am very sorry, but this would be wrong in this case. Because the number
>3.5 is not a constant known with (almost) no error, but a sample
>statistics based on the 1000 respondents. Therefore the
>Independent-samples T-test should be used as I've written in my previous
>posting.
>
>Jan
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
>John Antonakis
>Sent: Tuesday, December 19, 2006 9:25 AM
>To: [hidden email]
>Subject: Re: Significant difference - Means
>
>If you wiysh to test the mean of a particular group against the mean of a
>particular value then you can use the single-sample t-test (from the Menus
>Analyze/Compare Means/One sample t-test), or use the syntax:
>
>T-TEST
> /TESTVAL = 3.5
> /MISSING = ANALYSIS
> /VARIABLES = VAR00001
> /CRITERIA = CI(.95) .
>
>HTH,
>J.
>
>At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:
> >Hi to all,
> >
> >
> >
> >My first question is what test should I use if I need to assess the
> >significance of the mean of some subgroup in comparison the mean of the
> >total population. For example I have 1000 respondents that answered a
> >question on 7-point scale. Their mean/average is 3,5. I have 500
> >students among those 1000 respondents with mean/average 2,9. I want to
> >know if these values are significantly different.
> >What test should I use in SPSS?
> >
> >
> >
> >Thanks to all in advance.
> >
> >
> >
> >Samir
> >
> >
> >
> >GfK BH
> >
> >Centar za istrazivanje trzista i javnog mnijenja d.o.o.
> >
> >
> >
> >Samir Omeroviæ
> >
> >Researcher
> >
> >
> >
> >71 000 Sarajevo, Skenderija 44
> >
> >Bosna i Hercegovina
> >
> >
> >
> >
> >
> >Tel: +387 33 550 300
> >
> >Fax: +387 33 444 226
> >
> >www.gfk.ba
> >
> >www.gfk.com
> >
> >
> >
> >
> >
> >
> >
> >-----------------------------------------
> >
> >This e-mail (and any attachment/s) contains confidential and/or
> >privileged information. If you are not the intended recipient (or have
> >received this e-mail in error) please notify the sender immediately
> >and destroy this e-mail. Any unauthorized copying, disclosure or
> >distribution of the material in this e-mail is strictly forbidden.
> >
> >Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
> >osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
> >unistiti primljenu e-poruku.
> >Neautorizirano kopiranje,
> >objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
> >
> >---------------------------------------
> >
> >
>
>___________________________________
>
>Prof. John Antonakis
>Faculty of Management and Economics
>University of Lausanne
>Internef #527
>CH-1015 Lausanne-Dorigny
>Switzerland
>
>Tel: ++41 (0)21 692-3438
>Fax: ++41 (0)21 692-3305
>
>http://www.hec.unil.ch/jantonakis
>___________________________________

Samir Omerovic

Re: Significant difference - Means

In reply to this post by Spousta Jan

Hi Jan,

I think that I can not use Independent samples T-test since I have dependent variables. Those 500
students are the part of 1000 respondents. I do not want to test students vs. non-students but
students vs. all (non-students + students).

Samir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Spousta Jan
Sent: Tuesday, December 19, 2006 9:31 AM
To: [hidden email]
Subject: Re: Significant difference - Means

I am very sorry, but this would be wrong in this case. Because the number 3.5 is not a constant
known with (almost) no error, but a sample statistics based on the 1000 respondents. Therefore the
Independent-samples T-test should be used as I've written in my previous posting.

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John Antonakis
Sent: Tuesday, December 19, 2006 9:25 AM
To: [hidden email]
Subject: Re: Significant difference - Means

If you wiysh to test the mean of a particular group against the mean of a particular value then you
can use the single-sample t-test (from the Menus Analyze/Compare Means/One sample t-test), or use
the syntax:

T-TEST
/TESTVAL = 3.5
/MISSING = ANALYSIS
/VARIABLES = VAR00001
/CRITERIA = CI(.95) .

HTH,
J.

At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:

Spousta Jan

Re: Significant difference - Means

In reply to this post by Samir Omerovic

Hi John, the error of that 3.5 is about sqrt(1/1000) = 0,03 while the error of 2.9 for students is about sqrt(1/500) = 0.045. That is both errors are of the same order of magnitude and the population error cannot be neglected in this case.

Merry Christmas,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John Antonakis
Sent: Tuesday, December 19, 2006 9:46 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes that 3.5 represents the population mean (with a sample of 1,000 the standard error should be very small) then that is precisely the case in which one would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that the independent sample t-test is more appropriate (because it takes into account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

>I am very sorry, but this would be wrong in this case. Because the
>number
>3.5 is not a constant known with (almost) no error, but a sample
>statistics based on the 1000 respondents. Therefore the
>Independent-samples T-test should be used as I've written in my
>previous posting.
>
>Jan
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf
>Of John Antonakis
>Sent: Tuesday, December 19, 2006 9:25 AM
>To: [hidden email]
>Subject: Re: Significant difference - Means
>
>If you wiysh to test the mean of a particular group against the mean of
>a particular value then you can use the single-sample t-test (from the
>Menus Analyze/Compare Means/One sample t-test), or use the syntax:
>
>T-TEST
> /TESTVAL = 3.5
> /MISSING = ANALYSIS
> /VARIABLES = VAR00001
> /CRITERIA = CI(.95) .
>
>HTH,
>J.
>
>At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:
> >Hi to all,
> >
> >
> >
> >My first question is what test should I use if I need to assess the
> >significance of the mean of some subgroup in comparison the mean of
> >the total population. For example I have 1000 respondents that
> >answered a question on 7-point scale. Their mean/average is 3,5. I
> >have 500 students among those 1000 respondents with mean/average 2,9.
> >I want to know if these values are significantly different.
> >What test should I use in SPSS?
> >
> >
> >
> >Thanks to all in advance.
> >
> >
> >
> >Samir
> >
> >
> >
> >GfK BH
> >
> >Centar za istrazivanje trzista i javnog mnijenja d.o.o.
> >
> >
> >
> >Samir Omeroviæ
> >
> >Researcher
> >
> >
> >
> >71 000 Sarajevo, Skenderija 44
> >
> >Bosna i Hercegovina
> >
> >
> >
> >
> >
> >Tel: +387 33 550 300
> >
> >Fax: +387 33 444 226
> >
> >www.gfk.ba
> >
> >www.gfk.com
> >
> >
> >
> >
> >
> >
> >
> >-----------------------------------------
> >
> >This e-mail (and any attachment/s) contains confidential and/or
> >privileged information. If you are not the intended recipient (or
> >have received this e-mail in error) please notify the sender
> >immediately and destroy this e-mail. Any unauthorized copying,
> >disclosure or distribution of the material in this e-mail is strictly forbidden.
> >
> >Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
> >osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
> >unistiti primljenu e-poruku.
> >Neautorizirano kopiranje,
> >objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
> >
> >---------------------------------------
> >
> >
>
>___________________________________
>
>Prof. John Antonakis
>Faculty of Management and Economics
>University of Lausanne
>Internef #527
>CH-1015 Lausanne-Dorigny
>Switzerland
>
>Tel: ++41 (0)21 692-3438
>Fax: ++41 (0)21 692-3305
>
>http://www.hec.unil.ch/jantonakis
>___________________________________

statisticsdoc

Re: Significant difference - Means

Hi Jan and John,

Jan is correct. With the information that Samir possesses, formally one
should test the null hypothesis that the two samples have the same mean,
which is accomplished by using the independent groups t-test. Samir does
not possess information about the population mean. The mean of a sample of
1,000 is merely an estimate of the population mean - it is a statistic, not
a parameter.

Regards to All,

Stephen Brand

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Spousta Jan
Sent: Tuesday, December 19, 2006 3:54 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi John, the error of that 3.5 is about sqrt(1/1000) = 0,03 while the error
of 2.9 for students is about sqrt(1/500) = 0.045. That is both errors are of
the same order of magnitude and the population error cannot be neglected in
this case.

Merry Christmas,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John Antonakis
Sent: Tuesday, December 19, 2006 9:46 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes
that 3.5 represents the population mean (with a sample of 1,000 the standard
error should be very small) then that is precisely the case in which one
would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that
the independent sample t-test is more appropriate (because it takes into
account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

forbidden.

> >
> >Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
> >osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
> >unistiti primljenu e-poruku.
> >Neautorizirano kopiranje,
> >objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
> >
> >---------------------------------------
> >
> >
>
>___________________________________
>
>Prof. John Antonakis
>Faculty of Management and Economics
>University of Lausanne
>Internef #527
>CH-1015 Lausanne-Dorigny
>Switzerland
>
>Tel: ++41 (0)21 692-3438
>Fax: ++41 (0)21 692-3305
>
>http://www.hec.unil.ch/jantonakis
>___________________________________

Arthur Kramer

Re: Significant difference - Means

If my two cents is worth anything...

We are assuming the population is 1000 (i.e., N=1000) and the sample is 500
(n=500), and the population mean is known (3.5). Do we also know the
population standard deviation? If so, then the correct test is a z-test.
If the population standard deviation is not known, the correct test is the
single sample t-test. Again, that is, if the N=1000 is assumed a
population. However, with a sample of that size, i.e., 500, just about any
mean difference will obtain statistical significance. Samir would do better
to find an effect size, like Cohen's d, to test the magnitude of effect of
pulling out that sample from the population.

Arthur Kramer

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Statisticsdoc
Sent: Tuesday, December 19, 2006 8:16 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan and John,

Jan is correct. With the information that Samir possesses, formally one
should test the null hypothesis that the two samples have the same mean,
which is accomplished by using the independent groups t-test. Samir does
not possess information about the population mean. The mean of a sample of
1,000 is merely an estimate of the population mean - it is a statistic, not
a parameter.

Regards to All,

Stephen Brand

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Spousta Jan
Sent: Tuesday, December 19, 2006 3:54 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi John, the error of that 3.5 is about sqrt(1/1000) = 0,03 while the error
of 2.9 for students is about sqrt(1/500) = 0.045. That is both errors are of
the same order of magnitude and the population error cannot be neglected in
this case.

Merry Christmas,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John Antonakis
Sent: Tuesday, December 19, 2006 9:46 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes
that 3.5 represents the population mean (with a sample of 1,000 the standard
error should be very small) then that is precisely the case in which one
would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that
the independent sample t-test is more appropriate (because it takes into
account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

forbidden.

Samir Omerovic

Re: Significant difference - Means

In reply to this post by statisticsdoc

Hi Stephen,

I think I manage to bring some confusion with my English. I will try to correct it now.
I have 1000 respondents (this is also my population). MEAN value for these 1000 respondents is X,X.
Subgroup made of 500 students has MEAN value Y,Y. I want to see if these two MEANS are significantly
different, that is if the students themselves significantly differ from the population (1000
respondents) which they are part of.

I was thinking about Paired sample T-Test but I am not sure if this is the right option.

Any help would be greatly appreciated

Samir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Statisticsdoc
Sent: Tuesday, December 19, 2006 2:16 PM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan and John,

Jan is correct. With the information that Samir possesses, formally one
should test the null hypothesis that the two samples have the same mean,
which is accomplished by using the independent groups t-test. Samir does
not possess information about the population mean. The mean of a sample of
1,000 is merely an estimate of the population mean - it is a statistic, not
a parameter.

Regards to All,

Stephen Brand

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Spousta Jan
Sent: Tuesday, December 19, 2006 3:54 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi John, the error of that 3.5 is about sqrt(1/1000) = 0,03 while the error
of 2.9 for students is about sqrt(1/500) = 0.045. That is both errors are of
the same order of magnitude and the population error cannot be neglected in
this case.

Merry Christmas,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John Antonakis
Sent: Tuesday, December 19, 2006 9:46 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes
that 3.5 represents the population mean (with a sample of 1,000 the standard
error should be very small) then that is precisely the case in which one
would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that
the independent sample t-test is more appropriate (because it takes into
account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

forbidden.

statisticsdoc

Re: Significant difference - Means

Hi Samir,

If your population consists of the 1000 students, then the mean of 3.5 is a
population parameter, and you would be justified is using the one-sample
t-test suggested by John.

Best,

Stephen Brand

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Samir Omerović
Sent: Tuesday, December 19, 2006 9:34 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Stephen,

I think I manage to bring some confusion with my English. I will try to
correct it now.
I have 1000 respondents (this is also my population). MEAN value for these
1000 respondents is X,X.
Subgroup made of 500 students has MEAN value Y,Y. I want to see if these two
MEANS are significantly
different, that is if the students themselves significantly differ from the
population (1000
respondents) which they are part of.

I was thinking about Paired sample T-Test but I am not sure if this is the
right option.

Any help would be greatly appreciated

Samir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Statisticsdoc
Sent: Tuesday, December 19, 2006 2:16 PM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan and John,

Jan is correct. With the information that Samir possesses, formally one
should test the null hypothesis that the two samples have the same mean,
which is accomplished by using the independent groups t-test. Samir does
not possess information about the population mean. The mean of a sample of
1,000 is merely an estimate of the population mean - it is a statistic, not
a parameter.

Regards to All,

Stephen Brand

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Spousta Jan
Sent: Tuesday, December 19, 2006 3:54 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi John, the error of that 3.5 is about sqrt(1/1000) = 0,03 while the error
of 2.9 for students is about sqrt(1/500) = 0.045. That is both errors are of
the same order of magnitude and the population error cannot be neglected in
this case.

Merry Christmas,

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
John Antonakis
Sent: Tuesday, December 19, 2006 9:46 AM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Jan:

I agree that the 3.5 is not a perfect measure. However, If Samir believes
that 3.5 represents the population mean (with a sample of 1,000 the standard
error should be very small) then that is precisely the case in which one
would use the single-sample test.

If he just wants to compare the means of the two groups then I agree that
the independent sample t-test is more appropriate (because it takes into
account the standard error of the population mean).

Regards,
John

At 09:30 19.12.2006 +0100, Spousta Jan wrote:

forbidden.

Richard Ristow

Re: Significant difference - Means

In reply to this post by Spousta Jan

To weigh in with two comments:

At 03:54 AM 12/19/2006, Spousta Jan wrote:

>The error of that 3.5 is about sqrt(1/1000) = 0,03 while the error of
>2.9 for students is about sqrt(1/500) = 0.045. That is both errors are
>of the same order of magnitude and the population error cannot be
>neglected in this case.

I'd like to second, and emphasize, this. Jan is clearly right here,
where the two groups are the same size. However, the same thing holds
when the sizes are quite different.

First, the t-test algorithm correctly allows for the increased
precision in measuring the mean in the larger group. Replacing it by a
constant only 'gains' you a little precision you don't really have.

Second, inequality of group size matters less than one might think.
Roughly, precision goes as the square root of sample size. (Under
'nice' conditions, that's exact: standard error of estimate goes as the
square root of sample size.) That means increasing the sample size
ten-fold leaves the SEE still 1/3 of the size it had - quite a long way
from letting it be considered a constant.

And at 10:42 AM 12/19/2006, Statisticsdoc (Stephen Brand) wrote:

>If your population consists of the 1000 students, then the mean of 3.5
>is a population parameter, and you would be justified is using the
>one-sample t-test suggested by John.

(This won't be quite fair to Stephen Brand, who'd also written
"Formally one should test the null hypothesis that the two samples have
the same mean, by using the independent groups t-test.")

There's a philosophical position, which I agree with, that will hardly
ever accept something like "[my] population consists of the 1000
students." The argument is that, even if those 1,000 students are all
you've ever seen or ever will see, their observed values constitute a
set generated by an underlying random mechanism, and that randomness
must be allowed for in estimation exactly as if you were aware of
100,000 similar students.

('Generated by an underlying random mechanism' is sometimes expressed
as 'drawn from a conceptually infinite population.' However, while this
is technically accurate, I don't blame anyone who considers a
'conceptually infinite population' a very odd notion.)

--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.15.25/593 - Release Date: 12/19/2006 1:17 PM

Arthur Kramer

Re: Significant difference - Means

So, Richard is advocating for performing an independent groups t-test and
using the pooled standard error as the denominator, because he says both
groups are samples.

I am wondering, though, by what methodology was the sample of 500 drawn from
the sample of 1000 (was the original N 1500, or is one half of the group
being compared to the other half of the group?), is it a random sample or
was there a selection criterion?

There is another philosophical perspective at play also, and that is the
scale of 1 to 7, standing as the only measure of the construct, is at best
and ordinal scale because my distance between 1 and 2 may differ from your
distance between 1 and 2.

Arthur Kramer

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Richard Ristow
Sent: Tuesday, December 19, 2006 2:59 PM
To: [hidden email]
Subject: Re: Significant difference - Means

To weigh in with two comments:

At 03:54 AM 12/19/2006, Spousta Jan wrote:

>The error of that 3.5 is about sqrt(1/1000) = 0,03 while the error of
>2.9 for students is about sqrt(1/500) = 0.045. That is both errors are
>of the same order of magnitude and the population error cannot be
>neglected in this case.

I'd like to second, and emphasize, this. Jan is clearly right here,
where the two groups are the same size. However, the same thing holds
when the sizes are quite different.

First, the t-test algorithm correctly allows for the increased
precision in measuring the mean in the larger group. Replacing it by a
constant only 'gains' you a little precision you don't really have.

Second, inequality of group size matters less than one might think.
Roughly, precision goes as the square root of sample size. (Under
'nice' conditions, that's exact: standard error of estimate goes as the
square root of sample size.) That means increasing the sample size
ten-fold leaves the SEE still 1/3 of the size it had - quite a long way
from letting it be considered a constant.

And at 10:42 AM 12/19/2006, Statisticsdoc (Stephen Brand) wrote:

>If your population consists of the 1000 students, then the mean of 3.5
>is a population parameter, and you would be justified is using the
>one-sample t-test suggested by John.

(This won't be quite fair to Stephen Brand, who'd also written
"Formally one should test the null hypothesis that the two samples have
the same mean, by using the independent groups t-test.")

There's a philosophical position, which I agree with, that will hardly
ever accept something like "[my] population consists of the 1000
students." The argument is that, even if those 1,000 students are all
you've ever seen or ever will see, their observed values constitute a
set generated by an underlying random mechanism, and that randomness
must be allowed for in estimation exactly as if you were aware of
100,000 similar students.

('Generated by an underlying random mechanism' is sometimes expressed
as 'drawn from a conceptually infinite population.' However, while this
is technically accurate, I don't blame anyone who considers a
'conceptually infinite population' a very odd notion.)

--
No virus found in this outgoing message.
Checked by AVG Free Edition.
Version: 7.5.432 / Virus Database: 268.15.25/593 - Release Date: 12/19/2006
1:17 PM

Art Kendall

Re: Significant difference - Means

In reply to this post by Samir Omerovic

In the ANOVA context there are either two exclusive groups of cases
measured on a variable (independent groups(samples) t-test
OR
There is one group of cases measured on one variable under two different
circumstances such as time. Repeated measures t-test.

Perhaps, it is a matter of translation? It sounds like you are saying
whether half the cases are different from all the cases on a variable.
In other words is a bowl with 500 apples and 500 oranges different from
the oranges in the same bowl.
This is a very unusual question, and as asked there are no conventional
statistical ways to approach an answer.

Art
Social Research Consultants

Samir Omerović wrote:

>Hi Jan,
>
>I think that I can not use Independent samples T-test since I have dependent variables. Those 500
>students are the part of 1000 respondents. I do not want to test students vs. non-students but
>students vs. all (non-students + students).
>
>Samir
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Spousta Jan
>Sent: Tuesday, December 19, 2006 9:31 AM
>To: [hidden email]
>Subject: Re: Significant difference - Means
>
>I am very sorry, but this would be wrong in this case. Because the number 3.5 is not a constant
>known with (almost) no error, but a sample statistics based on the 1000 respondents. Therefore the
>Independent-samples T-test should be used as I've written in my previous posting.
>
>Jan
>
>-----Original Message-----
>From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John Antonakis
>Sent: Tuesday, December 19, 2006 9:25 AM
>To: [hidden email]
>Subject: Re: Significant difference - Means
>
>If you wiysh to test the mean of a particular group against the mean of a particular value then you
>can use the single-sample t-test (from the Menus Analyze/Compare Means/One sample t-test), or use
>the syntax:
>
>T-TEST
> /TESTVAL = 3.5
> /MISSING = ANALYSIS
> /VARIABLES = VAR00001
> /CRITERIA = CI(.95) .
>
>HTH,
>J.
>
>At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:
>
>
>>Hi to all,
>>
>>
>>
>>My first question is what test should I use if I need to assess the
>>significance of the mean of some subgroup in comparison the mean of the
>>total population. For example I have 1000 respondents that answered a
>>question on 7-point scale. Their mean/average is 3,5. I have 500
>>students among those 1000 respondents with mean/average 2,9. I want to
>>know if these values are significantly different.
>>What test should I use in SPSS?
>>
>>
>>
>>Thanks to all in advance.
>>
>>
>>
>>Samir
>>
>>
>>
>>GfK BH
>>
>>Centar za istrazivanje trzista i javnog mnijenja d.o.o.
>>
>>
>>
>>Samir Omeroviæ
>>
>>Researcher
>>
>>
>>
>>71 000 Sarajevo, Skenderija 44
>>
>>Bosna i Hercegovina
>>
>>
>>
>>
>>
>>Tel: +387 33 550 300
>>
>>Fax: +387 33 444 226
>>
>>www.gfk.ba
>>
>>www.gfk.com
>>
>>
>>
>>
>>
>>
>>
>>-----------------------------------------
>>
>>This e-mail (and any attachment/s) contains confidential and/or
>>privileged information. If you are not the intended recipient (or have
>>received this e-mail in error) please notify the sender immediately
>>and destroy this e-mail. Any unauthorized copying, disclosure or
>>distribution of the material in this e-mail is strictly forbidden.
>>
>>Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
>>osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
>>unistiti primljenu e-poruku.
>>Neautorizirano kopiranje,
>>objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
>>
>>---------------------------------------
>>
>>
>>
>>
>
>___________________________________
>
>Prof. John Antonakis
>Faculty of Management and Economics
>University of Lausanne
>Internef #527
>CH-1015 Lausanne-Dorigny
>Switzerland
>
>Tel: ++41 (0)21 692-3438
>Fax: ++41 (0)21 692-3305
>
>http://www.hec.unil.ch/jantonakis
>___________________________________
>
>
>
>

Art Kendall
Social Research Consultants

Spousta Jan

Re: Significant difference - Means

In reply to this post by Samir Omerovic

Hi Samir,

I explained it already, but perhaps I was not able to put it clearly. So once again:

Your task to determine whether

Mean(students) = Mean(population)

is equivalent to the classical task to determine whether

Mean(students) = Mean(non-students)

which can be assessed using the classical T-test.

You find the proof in one of my earlier posts.

The bottom line: Test students against non-students (both samples are then independent). If sig < 0.05, then you know that Mean(students) differs from Mean(population) significantly; if sig > 0.05, you have no reason to think that Mean(students) differs from Mean(population), the difference is not significant at the 95% level.

Happy New Year,

Jan

-----Original Message-----
From: Samir Omerović [mailto:[hidden email]]
Sent: Tuesday, December 19, 2006 9:52 AM
To: Spousta Jan
Cc: [hidden email]
Subject: RE: Significant difference - Means

Hi Jan,

I think that I can not use Independent samples T-test since I have dependent variables. Those 500 students are the part of 1000 respondents. I do not want to test students vs. non-students but students vs. all (non-students + students).

Samir

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Spousta Jan
Sent: Tuesday, December 19, 2006 9:31 AM
To: [hidden email]
Subject: Re: Significant difference - Means

I am very sorry, but this would be wrong in this case. Because the number 3.5 is not a constant known with (almost) no error, but a sample statistics based on the 1000 respondents. Therefore the Independent-samples T-test should be used as I've written in my previous posting.

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John Antonakis
Sent: Tuesday, December 19, 2006 9:25 AM
To: [hidden email]
Subject: Re: Significant difference - Means

If you wiysh to test the mean of a particular group against the mean of a particular value then you can use the single-sample t-test (from the Menus Analyze/Compare Means/One sample t-test), or use the syntax:

T-TEST
/TESTVAL = 3.5
/MISSING = ANALYSIS
/VARIABLES = VAR00001
/CRITERIA = CI(.95) .

HTH,
J.

At 09:05 19.12.2006 +0100, =?iso-8859-2?Q?Samir_Omerovi=E6?= wrote:

>Hi to all,
>
>
>
>My first question is what test should I use if I need to assess the
>significance of the mean of some subgroup in comparison the mean of the
>total population. For example I have 1000 respondents that answered a
>question on 7-point scale. Their mean/average is 3,5. I have 500
>students among those 1000 respondents with mean/average 2,9. I want to
>know if these values are significantly different.
>What test should I use in SPSS?
>
>
>
>Thanks to all in advance.
>
>
>
>Samir
>
>
>
>GfK BH
>
>Centar za istrazivanje trzista i javnog mnijenja d.o.o.
>
>
>
>Samir Omeroviæ
>
>Researcher
>
>
>
>71 000 Sarajevo, Skenderija 44
>
>Bosna i Hercegovina
>
>
>
>
>
>Tel: +387 33 550 300
>
>Fax: +387 33 444 226
>
>www.gfk.ba
>
>www.gfk.com
>
>
>
>
>
>
>
>-----------------------------------------
>
>This e-mail (and any attachment/s) contains confidential and/or
>privileged information. If you are not the intended recipient (or have
>received this e-mail in error) please notify the sender immediately and
>destroy this e-mail. Any unauthorized copying, disclosure or
>distribution of the material in this e-mail is strictly forbidden.
>
>Ova e-poruka sadrzi povjerljive i/ili povlastene podatke. Ako niste
>osoba naznacena kao primalac molimo odmah upozoriti posiljaoca i
>unistiti primljenu e-poruku.
>Neautorizirano kopiranje,
>objavljivanje ili distribucija sadrzaja e-poruke striktno je zabranjeno.
>
>---------------------------------------
>
>

Richard Ristow

Re: Significant difference - Means

In reply to this post by Samir Omerovic

At 03:52 AM 12/19/2006, Samir Omeroviã wrote:

[Did you really write this the 19th of December?
It seems to have been distributed to SPSSX-L on
Friday 5 January, over two weeks later.]

>I do not want to test students vs. non-students
>but students vs. all (non-students + students).

We've discussed this as extensively as any topic
I remember. I count 42 postings to the list,
including yours: 17 under this subject head,
"Significant difference - Means"; and 25 under
subject head "Sample Means".

There's a clear consensus: you can't test
"students vs. all". At least, I haven't seen a
posting that suggests a way to, or where the
poster believes there is a way; and there are
some pretty good statisticians on this list.
Everybody (and I've posted saying this, myself)
says you need to re-cast the problem to "test
students vs. non-students". That goes for
parametric comparisons (t-test), ordinal
non-parametric comparisons, and categorical comparisons (chi-square).

I think I can see what you're thinking: you want
to know if students are 'atypical' of the general
population. But it's hard even to define that.
Even conceptually, it depends on the proportion
of students in the whole population.

- If students are a small fraction, the
'population' is pretty well represented by the
non-students, and testing students vs.
non-students will give you about what you want.

- If students are the majority of the population,
they define what the 'population' is like, and
they can't very well be 'atypical'.

- You have equal numbers, 500 of each, in your
sample. That likely is not representative of the
population, but if there are about the same
numbers of students and non-students in the
population, the question doesn't mean much. If
the students are 'atypical', so are the
non-students, and vice versa; neither represents
the 'population' better than the other does.

You might do roughly what you want by doing an
independent-samples t-test between students and
non-students, but halving the difference of group
means while keeping the group standard deviations
the same. (You'd probably have to do the
arithmetic by hand; I can't see a quick way in
SPSS. You may *NOT* do this by dividing all the
values by 2; that halves the group standard
deviations, which is not what you want.) This
test would have a lot less power - less ability
to find a difference - than the
independent-samples t-test between students and non-students.

Think hard: this may be what you've thought you
wanted, but not what serves your purposes best.
That's what many of us have kept saying to you.

(Jan Spousta has argued that determining whether

> Mean(students) = Mean(population)
>is equivalent to the classical task to determine whether
> Mean(students) = Mean(non-students)

If this is so, and Jan's a good thinker, then
"halving the difference of group means" isn't a good idea at all.)

-Best of luck,
Richard Ristow

Marta García-Granero

Re: Significant difference - Means

In reply to this post by Samir Omerovic

Hi Samir

I've seem to have missed all the fun while I was out...

After reading your original posting plus all the replies you got
(mainly those from Jan and Richard), I think the topic should be
closed about the correct statistical method you should use: 2 samples
independent t-test. Anyway, I'm going to add a small explanation. I
think you got a wrong idea about "dependent samples" concept, and that
might the root of the problem accepting Jan's advice:

You said:

SO> I think that I can not use Independent samples T-test since I
SO> have dependent variables. Those 500
SO> students are the part of 1000 respondents.

This is not correct. You can talk about dependent samples only if you
have MATCHED (paired) data. For instance:

Subject Before After
1 25 27
2 18 19
3 23 23
4 15 21
5 31 32
. . .
. . .
. . .

The above example is the most usual pairing of data (each subject is
measured twice, three times... over time, or data on left eye is
compared to right eye...). You can also have naturally matched data
(siblings) or artificially paired data, like in matched case-control
studies (1:1 or 1:2, 1:3 matched subjects on key features).

On the other hand, you talk about independent samples when the
subjects are different, and no effort has been made to match them (on
age, gender... or other attributes).

You also said
:
SO> I do not want to test students vs. non-students but
SO> students vs. all (non-students + students).

As you can see, the situation you are describing falls in neither
category, and you can't use any of the above approaches to answer that
question. Besides, the question could be considered statistically
wrong (you can't test part of a sample against the whole sample).

On the other hand, if you have SOLID theoretical reasons to consider
the 1000 students as a finite population, and not a sample drawn from
a bigger (technically infinite) population formed by all possible
students you could have studied, then you could use the 1000 students
mean as a parameter (not as a sample statistic) and run a one-sample
t-test, but not the one provided by SPSS (it assumes infinite
populations) but with a finite-population correction term (right now I
don't have the formula at hand, but Googling "finite smple correction"
showed a lot of promising links, like this one at Steve Simon's great
web site: http://www.childrens-mercy.org/stats/size/population.asp or
this one:
http://myphliputil.pearsoncmg.com/student/bp_berenson_bbs_9/section7_3.pdf
).

I insist, you must have solid reasons for considering your 1000
students as a limited population (there won't ever be more of them or
something like that?) and run a one-sample t-test with finite
population correction. Personally, I'd stick to the
independent-samples t-test (500 vs 500 students).

--
Regards,
Dr. Marta Garcia-Granero,PhD mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

Roberts, Michael

Re: Significant difference - Means

Game, Set, and Match!!!

mike

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Marta Garcka-Granero
Sent: Friday, January 05, 2007 1:50 PM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Samir

I've seem to have missed all the fun while I was out...

After reading your original posting plus all the replies you got (mainly those from Jan and Richard), I think the topic should be closed about the correct statistical method you should use: 2 samples independent t-test. Anyway, I'm going to add a small explanation. I think you got a wrong idea about "dependent samples" concept, and that might the root of the problem accepting Jan's advice:

You said:

SO> I think that I can not use Independent samples T-test since I have
SO> dependent variables. Those 500 students are the part of 1000
SO> respondents.

This is not correct. You can talk about dependent samples only if you have MATCHED (paired) data. For instance:

Subject Before After
1 25 27
2 18 19
3 23 23
4 15 21
5 31 32
. . .
. . .
. . .

The above example is the most usual pairing of data (each subject is measured twice, three times... over time, or data on left eye is compared to right eye...). You can also have naturally matched data
(siblings) or artificially paired data, like in matched case-control studies (1:1 or 1:2, 1:3 matched subjects on key features).

On the other hand, you talk about independent samples when the subjects are different, and no effort has been made to match them (on age, gender... or other attributes).

You also said
:
SO> I do not want to test students vs. non-students but students vs. all
SO> (non-students + students).

As you can see, the situation you are describing falls in neither category, and you can't use any of the above approaches to answer that question. Besides, the question could be considered statistically wrong (you can't test part of a sample against the whole sample).

On the other hand, if you have SOLID theoretical reasons to consider the 1000 students as a finite population, and not a sample drawn from a bigger (technically infinite) population formed by all possible students you could have studied, then you could use the 1000 students mean as a parameter (not as a sample statistic) and run a one-sample t-test, but not the one provided by SPSS (it assumes infinite
populations) but with a finite-population correction term (right now I don't have the formula at hand, but Googling "finite smple correction"
showed a lot of promising links, like this one at Steve Simon's great web site: http://www.childrens-mercy.org/stats/size/population.asp or this one:
http://myphliputil.pearsoncmg.com/student/bp_berenson_bbs_9/section7_3.pdf
).

I insist, you must have solid reasons for considering your 1000 students as a limited population (there won't ever be more of them or something like that?) and run a one-sample t-test with finite population correction. Personally, I'd stick to the independent-samples t-test (500 vs 500 students).

--
Regards,
Dr. Marta Garcia-Granero,PhD mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does not understand. SPSS syntax guide cannot supply this knowledge, and it is certainly no substitute for the basic understanding of statistics and statistical thinking that is essential for the wise choice of methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

statisticsdoc

Re: Significant difference - Means

In reply to this post by Marta García-Granero

When in doubt, do what Marta says :)

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Marta Garcķa-Granero
Sent: Friday, January 05, 2007 1:50 PM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Samir

I've seem to have missed all the fun while I was out...

After reading your original posting plus all the replies you got
(mainly those from Jan and Richard), I think the topic should be
closed about the correct statistical method you should use: 2 samples
independent t-test. Anyway, I'm going to add a small explanation. I
think you got a wrong idea about "dependent samples" concept, and that
might the root of the problem accepting Jan's advice:

You said:

SO> I think that I can not use Independent samples T-test since I
SO> have dependent variables. Those 500
SO> students are the part of 1000 respondents.

This is not correct. You can talk about dependent samples only if you
have MATCHED (paired) data. For instance:

Subject Before After
1 25 27
2 18 19
3 23 23
4 15 21
5 31 32
. . .
. . .
. . .

The above example is the most usual pairing of data (each subject is
measured twice, three times... over time, or data on left eye is
compared to right eye...). You can also have naturally matched data
(siblings) or artificially paired data, like in matched case-control
studies (1:1 or 1:2, 1:3 matched subjects on key features).

On the other hand, you talk about independent samples when the
subjects are different, and no effort has been made to match them (on
age, gender... or other attributes).

You also said
:
SO> I do not want to test students vs. non-students but
SO> students vs. all (non-students + students).

As you can see, the situation you are describing falls in neither
category, and you can't use any of the above approaches to answer that
question. Besides, the question could be considered statistically
wrong (you can't test part of a sample against the whole sample).

On the other hand, if you have SOLID theoretical reasons to consider
the 1000 students as a finite population, and not a sample drawn from
a bigger (technically infinite) population formed by all possible
students you could have studied, then you could use the 1000 students
mean as a parameter (not as a sample statistic) and run a one-sample
t-test, but not the one provided by SPSS (it assumes infinite
populations) but with a finite-population correction term (right now I
don't have the formula at hand, but Googling "finite smple correction"
showed a lot of promising links, like this one at Steve Simon's great
web site: http://www.childrens-mercy.org/stats/size/population.asp or
this one:
http://myphliputil.pearsoncmg.com/student/bp_berenson_bbs_9/section7_3.pdf
).

I insist, you must have solid reasons for considering your 1000
students as a limited population (there won't ever be more of them or
something like that?) and run a one-sample t-test with finite
population correction. Personally, I'd stick to the
independent-samples t-test (500 vs 500 students).

--
Regards,
Dr. Marta Garcia-Granero,PhD mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)

Arthur Kramer

Re: Significant difference - Means

Because this issue has been alive for so long, would it be too much to ask
of Samir to provide the list with the result of his independent t-test, and
include the descriptive statistics (i.e., means, standard deviations and
standard errors)?

Arthur Kramer

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Statisticsdoc
Sent: Saturday, January 06, 2007 11:15 PM
To: [hidden email]
Subject: Re: Significant difference - Means

When in doubt, do what Marta says :)

For personalized and professional consultation in statistics and research
design, visit
www.statisticsdoc.com

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]]On Behalf Of
Marta Garcķa-Granero
Sent: Friday, January 05, 2007 1:50 PM
To: [hidden email]
Subject: Re: Significant difference - Means

Hi Samir

I've seem to have missed all the fun while I was out...

After reading your original posting plus all the replies you got
(mainly those from Jan and Richard), I think the topic should be
closed about the correct statistical method you should use: 2 samples
independent t-test. Anyway, I'm going to add a small explanation. I
think you got a wrong idea about "dependent samples" concept, and that
might the root of the problem accepting Jan's advice:

You said:

SO> I think that I can not use Independent samples T-test since I
SO> have dependent variables. Those 500
SO> students are the part of 1000 respondents.

This is not correct. You can talk about dependent samples only if you
have MATCHED (paired) data. For instance:

Subject Before After
1 25 27
2 18 19
3 23 23
4 15 21
5 31 32
. . .
. . .
. . .

The above example is the most usual pairing of data (each subject is
measured twice, three times... over time, or data on left eye is
compared to right eye...). You can also have naturally matched data
(siblings) or artificially paired data, like in matched case-control
studies (1:1 or 1:2, 1:3 matched subjects on key features).

On the other hand, you talk about independent samples when the
subjects are different, and no effort has been made to match them (on
age, gender... or other attributes).

You also said
:
SO> I do not want to test students vs. non-students but
SO> students vs. all (non-students + students).

As you can see, the situation you are describing falls in neither
category, and you can't use any of the above approaches to answer that
question. Besides, the question could be considered statistically
wrong (you can't test part of a sample against the whole sample).

On the other hand, if you have SOLID theoretical reasons to consider
the 1000 students as a finite population, and not a sample drawn from
a bigger (technically infinite) population formed by all possible
students you could have studied, then you could use the 1000 students
mean as a parameter (not as a sample statistic) and run a one-sample
t-test, but not the one provided by SPSS (it assumes infinite
populations) but with a finite-population correction term (right now I
don't have the formula at hand, but Googling "finite smple correction"
showed a lot of promising links, like this one at Steve Simon's great
web site: http://www.childrens-mercy.org/stats/size/population.asp or
this one:
http://myphliputil.pearsoncmg.com/student/bp_berenson_bbs_9/section7_3.pdf
).

I insist, you must have solid reasons for considering your 1000
students as a limited population (there won't ever be more of them or
something like that?) and run a one-sample t-test with finite
population correction. Personally, I'd stick to the
independent-samples t-test (500 vs 500 students).

--
Regards,
Dr. Marta Garcia-Granero,PhD mailto:[hidden email]
Statistician

---
"It is unwise to use a statistical procedure whose use one does
not understand. SPSS syntax guide cannot supply this knowledge, and it
is certainly no substitute for the basic understanding of statistics
and statistical thinking that is essential for the wise choice of
methods and the correct interpretation of their results".

(Adapted from WinPepi manual - I'm sure Joe Abrahmson will not mind)