Predictive Analytics in Data Where housing ...

[Dale]

All,

I ask for your thoughts.

I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !
 statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.

I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?

Dale


Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment
Member-at-Large Publications: Asso. Assessment & Research in Counseling
Asso. Editor Qualitative Research JCD
Professor: CPE (Faculty Appointment)
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497

Never play a thing the same way twice.   Louis Armstrong

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

[Bibel, Daniel (POL)]

This is certainly the approach that SPSS has when dealing with law enforcement - buy our (very expensive) software, throw everything in it, and stand back and let the magic happen.

It is depressing to hear that song from anybody, even worse to hear it from a company that certainly knows better.



Daniel Bibel
Massachusetts State Police
Crime Reporting Unit
978-451-3731 (voice)
978-451-3707 (fax)


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: Thursday, October 31, 2013 2:35 PM
To: [hidden email]
Subject: Predictive Analytics in Data Where housing ...

All,

I ask for your thoughts.

I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !
 statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.

I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?

Dale


Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment Member-at-Large Publications: Asso. Assessment & Research in Counseling Asso. Editor Qualitative Research JCD
Professor: CPE (Faculty Appointment)
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497

Never play a thing the same way twice.   Louis Armstrong

=====================
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]

Truly unfortunate. 
I believe that a more effective and honest approach would be to say that the software would make it much more easy to work effectively and efficiently with statisticians and social scientists.  It reduces the amount of drudgery, but it is not magic.


On 11/1/2013 10:37 AM, Bibel, Daniel (POL) [via SPSSX Discussion] wrote:

This is certainly the approach that SPSS has when dealing with law enforcement - buy our (very expensive) software, throw everything in it, and stand back and let the magic happen.

It is depressing to hear that song from anybody, even worse to hear it from a company that certainly knows better.



Daniel Bibel
Massachusetts State Police
Crime Reporting Unit
978-451-3731 (voice)
978-451-3707 (fax)


-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: Thursday, October 31, 2013 2:35 PM
To: [hidden email]
Subject: Predictive Analytics in Data Where housing ...

All,

I ask for your thoughts.

I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !
 statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.

I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?

Dale


Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment Member-at-Large Publications: Asso. Assessment & Research in Counseling Asso. Editor Qualitative Research JCD
Professor: CPE (Faculty Appointment)
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497

Never play a thing the same way twice.   Louis Armstrong

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

---

If you reply to this email, your message will be added to the discussion below:

http://spssx-discussion.1045642.n5.nabble.com/Predictive-Analytics-in-Data-Where-housing-tp5722846p5722848.html

To start a new topic under SPSSX Discussion, email [hidden email]
To unsubscribe from SPSSX Discussion, click here.
NAML

[Dale]

On Fri, 1 Nov 2013 08:18:43 -0700, Art Kendall <[hidden email]> wrote:




It seems that there is a great deal of this type of thing hitting the
university administrators as well. Data mining software seems to be sending
the message that ease of use means no knowledge of the tools are needed.

Dale

>
>    Truly unfortunate.&nbsp;
>        I believe that a more effective and honest approach would be to
>        say that the software would make it much more easy to work
>        effectively and efficiently with statisticians and social
>        scientists.&nbsp; It reduces the amount of drudgery, but it is not
>        magic.
>
>
>
>      Art Kendall
>Social Research Consultants
>      On 11/1/2013 10:37 AM, Bibel, Daniel (POL) [via SPSSX Discussion]
>      wrote:
>
>     This is certainly the approach that SPSS has when
>      dealing with law enforcement - buy our (very expensive) software,
>      throw everything in it, and stand back and let the magic happen.
>
>
>      It is depressing to hear that song from anybody, even worse to
>      hear it from a company that certainly knows better.
>
>
>
>
>      Daniel Bibel
>
>      Massachusetts State Police
>
>      Crime Reporting Unit
>
>      978-451-3731 (voice)
>
>      978-451-3707 (fax)
>
>
>
>      -----Original Message-----
>
>      From: SPSSX(r) Discussion [mailto: [hidden email] ]
>      On Behalf Of Pietrzak, Dale
>
>      Sent: Thursday, October 31, 2013 2:35 PM
>
>      To: [hidden email]
>
>      Subject: Predictive Analytics in Data Where housing ...
>
>
>      All,
>
>
>      I ask for your thoughts.
>
>
>      I sat through a web presentation to our regental system put on by
>      IBM today. &nbsp;They are promoting their data where housing software
>      solutions. &nbsp;A part of that presentation was on predictive
>      analytics. &nbsp;These appeared to be essentially SPSS integrated
>      predictive modeling based largely in regression type models as one
>      might expect. &nbsp;The thing that surprised me was that the presenter
>      indicated that you did not need to understand statistics of be a
>      statistician in any way. &nbsp;He indicated one should dump all
>      variables into the model and the various models would find the
>      solution (which appeared to be variations in stepwise processes).
>      &nbsp;Further, that the statistical model you used really did not
>      depend on the data type but rather that like various dumps you
>      could allow the software to select the model for like with the
>      variables, you just let it run them all and the one you liked the
>      best for a solution was there for you to select. There was no
>      thought or the appropriateness of the !
>
>      &nbsp;statistical model based on the data types. I asked about the
>      potential for &nbsp;inappropriate use of such point-and-click
>      approaches and was essentially dismissed.
>
>
>      I realize data mining is hot right now, but is that &nbsp;really the
>      direction we are headed with analysis today? &nbsp;Is it really the
>      belief of the corporation that the software is all you need you do
>      not need any real expertise any longer?
>
>
>      Dale
>
>
>
>      Dale Pietrzak, Ed.D., LPCMH, CCMHC
>
>      Director, Office of Academic Evaluation and Assessment
>      Member-at-Large Publications: Asso. Assessment &amp; Research in
>      Counseling Asso. Editor Qualitative Research JCD
>
>      Professor: CPE (Faculty Appointment)
>
>      University of South Dakota
>
>      Slagle Hall Room 102
>
>      414 East Clark Street
>
>      605-677-6497
>
>
>      Never play a thing the same way twice. &nbsp; Louis Armstrong
>
>
>      =====================
>
>      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
>
>
>
>
>
>        If you reply to this email, your
>          message will be added to the discussion below:
>

http://spssx-discussion.1045642.n5.nabble.com/Predictive-Analytics-in-Data-Where-housing-tp5722846p5722848.html

>
>
>        To start a new topic under SPSSX Discussion, email
>        [hidden email]
>        To unsubscribe from SPSSX Discussion, click
>          here .
>        NAML
>
>
>
>
>
>
>
>
>-----
>Art Kendall
>Social Research Consultants
>--
>View this message in context:

http://spssx-discussion.1045642.n5.nabble.com/Predictive-Analytics-in-Data-Where-housing-tp5722846p5722850.html
>Sent from the SPSSX Discussion mailing list archive at Nabble.com.

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

[John F Hall]

My site IT wizard, Terry Blom, once worked for IBM.  He says they were only ever interested in $$$, £££ and €€€.

 

I remember when the "new" SPSS was launched in the UK at Heathrow Airport Novotel more than 30 years ago.  I was there with Jim Ring and we were horrified at the sales pitch for marketing rather than social (survey) research. Even more embarrassing was Major Lester attempting, and failing, to demonstrate SPSS in action on-screen. Jim and I had to intervene and explain (in whispered prompts) how to use SPSS.  Luckily we were sitting on the front row, close to the stage, so maybe the people at the back didn't notice.

 

Looks like they're at it again.

 

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email]   

Website: www.surveyresearch.weebly.com  

SPSS start page:  www.surveyresearch.weebly.com/spss-without-tears.html  

  

  

 

 

 

 

 

 

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: 31 October 2013 19:35
To: [hidden email]
Subject: Predictive Analytics in Data Where housing ...

 

All,

 

I ask for your thoughts.

 

I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !

statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.

 

I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?

 

Dale

 

 

Dale Pietrzak, Ed.D., LPCMH, CCMHC

Director, Office of Academic Evaluation and Assessment Member-at-Large Publications: Asso. Assessment & Research in Counseling Asso. Editor Qualitative Research JCD

Professor: CPE (Faculty Appointment)

University of South Dakota

Slagle Hall Room 102

414 East Clark Street

605-677-6497

 

Never play a thing the same way twice.   Louis Armstrong

 

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

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

[Bruce Weaver]

As John's post suggests, this sort of thing predates the IBM takeover of SPSS.  Does anyone else remember the "Real Stats. Real Easy."  slogan?  

(If you search the archives, you can find a post in which David expressed some mild disagreement with that approach.  I think his exact words were, "What a crock!")  ;-)

p.s. - Was that meant to be warehousing, by any chance?  :-)

John F Hall wrote

My site IT wizard, Terry Blom, once worked for IBM.  He says they were only ever interested in $$$, £££ and €€€.
 
I remember when the "new" SPSS was launched in the UK at Heathrow Airport Novotel more than 30 years ago.  I was there with Jim Ring and we were horrified at the sales pitch for marketing rather than social (survey) research. Even more embarrassing was Major Lester attempting, and failing, to demonstrate SPSS in action on-screen. Jim and I had to intervene and explain (in whispered prompts) how to use SPSS.  Luckily we were sitting on the front row, close to the stage, so maybe the people at the back didn't notice.
 
Looks like they're at it again.
 
 
John F Hall (Mr)
[Retired academic survey researcher]
 
Email:   [hidden email]   
Website: www.surveyresearch.weebly.com  
SPSS start page:  www.surveyresearch.weebly.com/spss-without-tears.html  
 
 
 
 
 
 
 
 
-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: 31 October 2013 19:35
To: [hidden email]
Subject: Predictive Analytics in Data Where housing ...
 
All,
 
I ask for your thoughts.
 
I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !
statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.
 
I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?
 
Dale
 
 
Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment Member-at-Large Publications: Asso. Assessment & Research in Counseling Asso. Editor Qualitative Research JCD
Professor: CPE (Faculty Appointment)
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497
 
Never play a thing the same way twice.   Louis Armstrong
 
=====================
To manage your subscription to SPSSX-L, send a message to  <mailto:[hidden email]> [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

[David Marso]

http://spssx-discussion.1045642.n5.nabble.com/Isolate-independent-variables-tp5714101p5714126.html
4th post in the thread ;-)
Maybe Dale was wavering between data warehousing and data whorehousing ;-(

Bruce Weaver wrote

As John's post suggests, this sort of thing predates the IBM takeover of SPSS.  Does anyone else remember the "Real Stats. Real Easy."  slogan?  

(If you search the archives, you can find a post in which David expressed some mild disagreement with that approach.  I think his exact words were, "What a crock!")  ;-)

p.s. - Was that meant to be warehousing, by any chance?  :-)

John F Hall wrote

My site IT wizard, Terry Blom, once worked for IBM.  He says they were only ever interested in $$$, £££ and €€€.
 
I remember when the "new" SPSS was launched in the UK at Heathrow Airport Novotel more than 30 years ago.  I was there with Jim Ring and we were horrified at the sales pitch for marketing rather than social (survey) research. Even more embarrassing was Major Lester attempting, and failing, to demonstrate SPSS in action on-screen. Jim and I had to intervene and explain (in whispered prompts) how to use SPSS.  Luckily we were sitting on the front row, close to the stage, so maybe the people at the back didn't notice.
 
Looks like they're at it again.
 
 
John F Hall (Mr)
[Retired academic survey researcher]
 
Email:   [hidden email]   
Website: www.surveyresearch.weebly.com  
SPSS start page:  www.surveyresearch.weebly.com/spss-without-tears.html  
 
 
 
 
 
 
 
 
-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Pietrzak, Dale
Sent: 31 October 2013 19:35
To: [hidden email]
Subject: Predictive Analytics in Data Where housing ...
 
All,
 
I ask for your thoughts.
 
I sat through a web presentation to our regental system put on by IBM today.  They are promoting their data where housing software solutions.  A part of that presentation was on predictive analytics.  These appeared to be essentially SPSS integrated predictive modeling based largely in regression type models as one might expect.  The thing that surprised me was that the presenter indicated that you did not need to understand statistics of be a statistician in any way.  He indicated one should dump all variables into the model and the various models would find the solution (which appeared to be variations in stepwise processes).  Further, that the statistical model you used really did not depend on the data type but rather that like various dumps you could allow the software to select the model for like with the variables, you just let it run them all and the one you liked the best for a solution was there for you to select. There was no thought or the appropriateness of the !
statistical model based on the data types. I asked about the potential for  inappropriate use of such point-and-click approaches and was essentially dismissed.
 
I realize data mining is hot right now, but is that  really the direction we are headed with analysis today?  Is it really the belief of the corporation that the software is all you need you do not need any real expertise any longer?
 
Dale
 
 
Dale Pietrzak, Ed.D., LPCMH, CCMHC
Director, Office of Academic Evaluation and Assessment Member-at-Large Publications: Asso. Assessment & Research in Counseling Asso. Editor Qualitative Research JCD
Professor: CPE (Faculty Appointment)
University of South Dakota
Slagle Hall Room 102
414 East Clark Street
605-677-6497
 
Never play a thing the same way twice.   Louis Armstrong
 
=====================
To manage your subscription to SPSSX-L, send a message to  <mailto:[hidden email]> [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