Statistics and signal detection

[Frederic Villamayor Forcada]

Dear co-listers:

I know this is not really an SPSS question, but I hope I will find help
here. In any case, any answer can be submitted to me directly, not to the
list.
I'm preparing a presentation on "Sample size calculation" for a
non-statistician audience.
One of my ideas has been to illustrate the process of statistical testing
by comparing it with that of correctly dialing a radio station: I've
assumed that the background noise could be the "variability of the
population", the intensity of the signal could be the"effect size", the
radio receptor could be the "experimental design", and the difficulty to
correctly sintonize the radio station could be the "sample size". It will
be easier to sintonize a station if the signal is good, if there is low
background noise and the radio receptor has a reasonabli quality.
However, I'm blocked trying to imagine what would be the "statistical
error" in this example.  Can anybody help me? Or it is really an absurd
idea?

Thank you in advance.

Frederic


Frederic Villamayor, BS
Researcher
Biostatistics Unit
Juan de Sada, 32
08028 Barcelona
Tel    [phone]
Fax   [phone]
[hidden email]
http://www.ferrergrupo.com 

Este mensaje, y en su caso, cualquier fichero anexo al mismo, puede
contener información confidencial, siendo para uso exclusivo del
destinatario, quedando prohibida su divulgación, copia o distribución a
terceros sin la autorización expresa del remitente. Si Vd. ha recibido
este mensaje erróneamente, se ruega lo notifique al remitente y proceda a
su borrado. Gracias por su colaboración.

This message and its annexed files may contain confidential information
which is exclusively for the use of the addressee. It is strictly
forbidden to distribute copies to third parties without the explicit
permission of the sender. If you receive this message by mistake, please
notify it to the sender and make sure to delete it. Thank you for your
kind cooperation.

[Spousta Jan]

Dear Frederic,

It would be perhaps the noise you still hear from the receiver. - But the question is whether the metapher is not about as complicated as the matter it should explain. :-)

Best regards

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Frederic Villamayor Forcada
Sent: Monday, June 11, 2007 12:54 PM
To: [hidden email]
Subject: Statistics and signal detection

Dear co-listers:

I know this is not really an SPSS question, but I hope I will find help here. In any case, any answer can be submitted to me directly, not to the list.
I'm preparing a presentation on "Sample size calculation" for a non-statistician audience.
One of my ideas has been to illustrate the process of statistical testing by comparing it with that of correctly dialing a radio station: I've assumed that the background noise could be the "variability of the population", the intensity of the signal could be the"effect size", the radio receptor could be the "experimental design", and the difficulty to correctly sintonize the radio station could be the "sample size". It will be easier to sintonize a station if the signal is good, if there is low background noise and the radio receptor has a reasonabli quality.
However, I'm blocked trying to imagine what would be the "statistical error" in this example.  Can anybody help me? Or it is really an absurd idea?

Thank you in advance.

Frederic


Frederic Villamayor, BS
Researcher
Biostatistics Unit
Juan de Sada, 32
08028 Barcelona
Tel    [phone]
Fax   [phone]
[hidden email]
http://www.ferrergrupo.com

Este mensaje, y en su caso, cualquier fichero anexo al mismo, puede contener información confidencial, siendo para uso exclusivo del destinatario, quedando prohibida su divulgación, copia o distribución a terceros sin la autorización expresa del remitente. Si Vd. ha recibido este mensaje erróneamente, se ruega lo notifique al remitente y proceda a su borrado. Gracias por su colaboración.

This message and its annexed files may contain confidential information which is exclusively for the use of the addressee. It is strictly forbidden to distribute copies to third parties without the explicit permission of the sender. If you receive this message by mistake, please notify it to the sender and make sure to delete it. Thank you for your kind cooperation.

[samir-9]

the assumed extent of tolerance in terms of difficulty in sintonizing the
radion station should be the sampling error (in your idea!).

----- Original Message -----
From: "Frederic Villamayor Forcada" <[hidden email]>
To: <[hidden email]>
Sent: 11 June, 2007 4:53 PM
Subject: Statistics and signal detection


Dear co-listers:

I know this is not really an SPSS question, but I hope I will find help
here. In any case, any answer can be submitted to me directly, not to the
list.
I'm preparing a presentation on "Sample size calculation" for a
non-statistician audience.
One of my ideas has been to illustrate the process of statistical testing
by comparing it with that of correctly dialing a radio station: I've
assumed that the background noise could be the "variability of the
population", the intensity of the signal could be the"effect size", the
radio receptor could be the "experimental design", and the difficulty to
correctly sintonize the radio station could be the "sample size". It will
be easier to sintonize a station if the signal is good, if there is low
background noise and the radio receptor has a reasonabli quality.
However, I'm blocked trying to imagine what would be the "statistical
error" in this example.  Can anybody help me? Or it is really an absurd
idea?

Thank you in advance.

Frederic


Frederic Villamayor, BS
Researcher
Biostatistics Unit
Juan de Sada, 32
08028 Barcelona
Tel    [phone]
Fax   [phone]
[hidden email]
http://www.ferrergrupo.com

Este mensaje, y en su caso, cualquier fichero anexo al mismo, puede
contener información confidencial, siendo para uso exclusivo del
destinatario, quedando prohibida su divulgación, copia o distribución a
terceros sin la autorización expresa del remitente. Si Vd. ha recibido
este mensaje erróneamente, se ruega lo notifique al remitente y proceda a
su borrado. Gracias por su colaboración.

This message and its annexed files may contain confidential information
which is exclusively for the use of the addressee. It is strictly
forbidden to distribute copies to third parties without the explicit
permission of the sender. If you receive this message by mistake, please
notify it to the sender and make sure to delete it. Thank you for your
kind cooperation.


--
No virus found in this incoming message.
Checked by AVG Free Edition.
Version: 7.5.472 / Virus Database: 269.8.9/834 - Release Date: 05-Jun-07
2:38 PM

[samir-9]

the metaphor should not be deemed as a complicated one if his audience are
radio engineers - they better understand radio transmission parlance than
statistics -> sampling -> sampling error.

----- Original Message -----
From: "Spousta Jan" <[hidden email]>
To: <[hidden email]>
Sent: 11 June, 2007 4:58 PM
Subject: Re: Statistics and signal detection

> Dear Frederic,
>
> It would be perhaps the noise you still hear from the receiver. - But the
> question is whether the metapher is not about as complicated as the matter
> it should explain. :-)
>
> Best regards
>
> Jan
>
> -----Original Message-----
> From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
> Frederic Villamayor Forcada
> Sent: Monday, June 11, 2007 12:54 PM
> To: [hidden email]
> Subject: Statistics and signal detection
>
> Dear co-listers:
>
> I know this is not really an SPSS question, but I hope I will find help
> here. In any case, any answer can be submitted to me directly, not to the
> list.
> I'm preparing a presentation on "Sample size calculation" for a
> non-statistician audience.
> One of my ideas has been to illustrate the process of statistical testing
> by comparing it with that of correctly dialing a radio station: I've
> assumed that the background noise could be the "variability of the
> population", the intensity of the signal could be the"effect size", the
> radio receptor could be the "experimental design", and the difficulty to
> correctly sintonize the radio station could be the "sample size". It will
> be easier to sintonize a station if the signal is good, if there is low
> background noise and the radio receptor has a reasonabli quality.
> However, I'm blocked trying to imagine what would be the "statistical
> error" in this example.  Can anybody help me? Or it is really an absurd
> idea?
>
> Thank you in advance.
>
> Frederic
>
>
> Frederic Villamayor, BS
> Researcher
> Biostatistics Unit
> Juan de Sada, 32
> 08028 Barcelona
> Tel    [phone]
> Fax   [phone]
> [hidden email]
> http://www.ferrergrupo.com
>
> Este mensaje, y en su caso, cualquier fichero anexo al mismo, puede
> contener información confidencial, siendo para uso exclusivo del
> destinatario, quedando prohibida su divulgación, copia o distribución a
> terceros sin la autorización expresa del remitente. Si Vd. ha recibido
> este mensaje erróneamente, se ruega lo notifique al remitente y proceda a
> su borrado. Gracias por su colaboración.
>
> This message and its annexed files may contain confidential information
> which is exclusively for the use of the addressee. It is strictly
> forbidden to distribute copies to third parties without the explicit
> permission of the sender. If you receive this message by mistake, please
> notify it to the sender and make sure to delete it. Thank you for your
> kind cooperation.
>
>
> --
> No virus found in this incoming message.
> Checked by AVG Free Edition.
> Version: 7.5.472 / Virus Database: 269.8.9/834 - Release Date: 05-Jun-07
> 2:38 PM
>

[Frommelt, Robert A.]

Does anyone know a way to calculate the confidence interval around the
median using SPSS?

Thanks,

Allen

CONFIDENTIALITY NOTICE:  This communication contains information
intended for the use of the individuals to whom it is addressed
and may contain information that is privileged, confidential or
exempt from other disclosure under applicable law.  If you are
not the intended recipient, you are notified that any disclosure,
printing, copying, distribution or use of the contents is prohibited.
If you have received this in error, please notify the sender
immediately by telephone or by returning it by return mail and then
permanently delete the communication from your system.  Thank you.

[David Hitchin]

> Does anyone know a way to calculate the confidence interval around
> the
> median using SPSS?

Bootstrapping (repeated resampling with replacement) is easy.

See

http://www.spsstools.net/
where in the Syntax section under "Bootstrap and random numbers", the
5th macro listed deals with mean and median.

Alternatively you could use the method described in
http://www-users.york.ac.uk/~mb55/intro/cicent.htm

You will find references to both methods in
http://www.umanitoba.ca/centres/mchp/concept/dict/ci_median/

David Hitchin

[Marta Garcia-Granero]

Hi everybody

This is Marta (usually [hidden email]), writing from a different
mail because my ADSL is dead and technical assistance is taking a lot of
time to take a look at it. Anyway, my brand new laptop has a wireless card,
and some kind neighbour has an absolutely unprotected wireless connection I
can access from one corner of my house. Since he/she is not using it right
now, I can use it without disturbing him/her.

Now, my reply

Although you can use boostrapping mentods to compute a 95%CI for the median,
you can also use RATIO procedure (available since SPSS 11) to compute it.
You just need a little trick:

TEMPORARY.
COMPUTE k=1.
RATIO STATISTICS "yourvariable" WITH k
   /PRINT =  CIN(95) MEDIAN .

If you want to compute CI for the median of the ifferences between two
samples, then you need a different approach (see Altman's book "Statistics
with confidence"). I have a couple of macros (for paired and unpaired
samples)
.

If you want them, please give me a couple of days to get back to normal
Internet connection (and to transport all my files from the old to the new
computer) and I'll send them.

Remember to access me thru the list, not privately, please.

Regards,
Marta, the sporadic bandwidth stealer, ;)