Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

[John F Hall]

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [mailto:[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

[Thomas MacFarland]

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language.

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

[John F Hall]

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

[Jason Lamprianou]

Dear all, 

because of the well-known crises in the European counties around the Mediterranean sea, many governmental departments have lost their SPSS licenses. My wive's department, in Cyprus, lost the SPSS lisences, and she had to urgently learn R so that she would teach it to others. Life is forcing more and more people everyday towards R or other free software. Unless the commercial giants like SPSS reduce their prices, they will loose more ground because the competition is getting better and it is FREE.

Jason

On 22 January 2014 19:33, John F Hall <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:[hidden email]]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice <a href="tel:[phone]" value="[phone]" target="_blank">954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

--

Iasonas Lamprianou

Department of Social and Political Sciences

University of Cyprus

[Mike Griffiths-2]

People might be interested to know that there is a free rip-off version of SPSS called PSPP:  http://www.gnu.org/software/pspp/.  They are currently  calling it version 0.8.2, and the initial zero suggests to me that it might not yet be fully proven, but perhaps I'm just paranoid since it doesn't say that on the home page.  When I tried it out a couple of years ago I found a silly error which I reported and they have fixed, but I haven't really tried it out since, because I don't need to.

 

Mike Griffiths

---

From: [hidden email] <[hidden email]> on behalf of John F Hall <[hidden email]>
Sent: 22 January 2014 17:33
To: 'Dr. Thomas W. MacFarland'; [hidden email]; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

[Jacqueline Carter]

Thanks to all for this discussion. 

I'd be tempted to forward this to the IASSIST list as they will have experience with R in teaching - also ICPSR. I'll forward their responses to this list. 

I also know that @cogdog - Alan Levine at Leicester - has been using R with his students (not social scientists) so will ask for his comments. He and I talked previously about development of reusable resources for teaching with R. 

Best,

Jackie 

Sent from my iPhone

Dr Jackie Carter

UK Data Service: Director for Communications and Impact

University of Manchester Q-Step Co-director 

Mimas Senior Manager 

07747460963

@JackieCarter

On 22 Jan 2014, at 17:36, "John F Hall" <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

[Yang, Hongwei]

Dear all,

 

I use R, SPSS, and SAS on a regular basis. I like all three programs and think each one has its own strengths and weaknesses.

 

For R: Many people are switching to R, which is true. However, since R is free, its quality control is not as good as commercial programs like SPSS and SAS: We do not know how good the estimates are. There is no technical support. Also, it is not quite easy to learn.

 

For SPSS: It is easy to use and produces estimates that are trustworthy. I just hope it can add more advanced modeling stuff as built-in routines, like what STATA does.

 

Hongwei Yang, University of Kentucky

 

 

 

From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of Jason Lamprianou
Sent: Wednesday, January 22, 2014 1:00 PM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Dear all, 

because of the well-known crises in the European counties around the Mediterranean sea, many governmental departments have lost their SPSS licenses. My wive's department, in Cyprus, lost the SPSS lisences, and she had to urgently learn R so that she would teach it to others. Life is forcing more and more people everyday towards R or other free software. Unless the commercial giants like SPSS reduce their prices, they will loose more ground because the competition is getting better and it is FREE.

 

Jason

 

On 22 January 2014 19:33, John F Hall <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:[hidden email]]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5" target="_blank">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice <a href="tel:[phone]" target="_blank">954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

 

--

 

 

Iasonas Lamprianou

 

Department of Social and Political Sciences

University of Cyprus

[Thomas MacFarland]

John and Everyone:

First, the SPSS video that I prepared a few years ago was moved to a new machine and it is now available at https://tegr.it/y/u9te.  Again, it is not password protected but unfortunately the video is a bit old and I need to update it, but the basics on SPSS are still the same – just a minor change here and there on the interface.

 

I do not want to get into a discussion on the virtues of one software program over another.  R and SPSS are both great and I enjoy using them.

 

However, I have two concerns about SPSS that must be considered:

 

The SPSS license is expensive and with our current budget challenges (at least for colleges and universities in the US) it is only prudent to have a Plan B in case the suits in accounting with sharp pencils and green eyeshades tell the Deans and VPs to cut their budgets by X percent.  Software licenses are always an easy target for cuts.

 

Similar to many other technology companies, IBM has their own challenges now given current movement in stock price and pressure on IBM puts pressure on future iterations of SPSS.

 

So – I hope to use SPSS for many years to come, but I also plan to use R for a few select features absent from SPSS and to also have some degree of insurance against the future.

 

Best wishes.

 

Tom MacFarland 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: John F Hall [mailto:[hidden email]]
Sent: Wednesday, January 22, 2014 12:34 PM
To: Dr. Thomas W. MacFarland; [hidden email]; [hidden email]
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

[John F Hall]

Nicole

 

This needs to be fed to the SPSS list as well (no embedded links or attachments allowed). 

 

Social science students are interested in social questions, not formulae or programming languages.  SPSS syntax is far more like English language and is far quicker and easier for basic operations.  In any case, as my old boss the late Dr Mark Abrams used to say, “If it’s worth saying, you can say it in percentages.” although I suspect he would accept some of the graphic output from SPSS as an alternative.

 

John

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of Nicole Janz
Sent: 22 January 2014 20:02
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Dear all,

 

the Cambridge Social Sciences Research Methods Centre (SSRMC) has changed most basic stats courses to R last academic year. I was the one converting the courses from SPSS to R (exercises, exams, lectures) and we did a 'trial' to see how it goes: A colleague taught the same courses in SPSS, while I taught in R. We did this with Bivariate Statistics and Regression. From our experience, which includes class interaction, student evaluations and feedback as well as exam results, there was no difference for the majority of students! We now teach all basic courses in R since students can use their own laptops in class. Students have accepted it. (More specifically, we use Rstudio using a prepared Rscript in class.) We also switched the main textbook from Andy Field's SPSS version to Field's R version.

 

I do not want to pass judgement on SPSS since I only used it for a year or so, but I'd say there is no disadvantage for a stats beginner learning R, and if they students want to go on to more flexible modelling and data handling later they have a head start.  

 

I do see though that anyone who has worked for years in another software which is click-based, a change might sound unnecessary - especially if all materials are already prepared and well documented. It took me 4-5 days to convert each 2-hour lecture from SPSS to Rscript, and to make sure it all works.

 

If anyone wants to have a look and download materials for basic statistics for social scientists, incl. exercises and datasets in R, I make that available here: www.nicolejanz.de/teaching.html.

 

I think there's room for all kinds of software - I'm just joining the discussion to show that a switch to R is not as frightening as some think (neither for instructors nor students).

 

Best,

Nicole

 

 

 

 

 

Nicole Janz, PhD Cand.

Lecturer at Social Sciences Research Methods Centre

University of Cambridge

Department of Politics and International Studies

www.nicolejanz.de | [hidden email] | 

Skype: nicole.janz

Blog: politicalsciencereplication.wordpress.com

Twitter: @PolSciReplicate

 

 

 

 

 

 

 

 

 

 

 

 

On Jan 22, 2014, at 6:11 PM, Jacqueline Carter wrote:

Thanks to all for this discussion. 

 

I'd be tempted to forward this to the IASSIST list as they will have experience with R in teaching - also ICPSR. I'll forward their responses to this list. 

 

I also know that @cogdog - Alan Levine at Leicester - has been using R with his students (not social scientists) so will ask for his comments. He and I talked previously about development of reusable resources for teaching with R. 

 

Best,

Jackie 

Sent from my iPhone

 

Dr Jackie Carter

UK Data Service: Director for Communications and Impact

University of Manchester Q-Step Co-director 

Mimas Senior Manager 

 

07747460963

@JackieCarter

On 22 Jan 2014, at 17:36, "John F Hall" <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email]  

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

 

[John F Hall]

 

 

From: Reades, Jonathan [mailto:[hidden email]]
Sent: 22 January 2014 19:47
To: John F Hall
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

 

On 22 Jan 2014, at 17:33, John F Hall <[hidden email]> wrote:

<snip>

SPSS is used by millions worldwide: how many use R?  

 

Although I agree with most of what you've said below, I'm not sure that this first point is the strongest -- R is being used by everyone from Google to pharma companies to GIS researchers, so in terms of having access to libraries that do interesting (read: powerful, cutting edge) things R is miles in advance of SPSS already and accelerating. I could cheekily put this as: Windows XP is used by millions worldwide, does that mean it's what we should be teaching our students?

 

The last GIS conference I went to, the majority of the new research was being presented had been done in R and Python, not ArcGIS. So there are *already* spatial analysis libraries available in R that aren't yet, and may well not ever be, available for the current 'industry standard' platform. So in one application I can do everything from machine learning to basic statistics all using the same syntax/language. I would agree 100% that the hurdles are *definitely* higher, but the payoff is greater.

 

Of course, that payoff doesn't necessarily make it an appropriate teaching tool for social sciences students, but as one of the other respondents noted, the fact that it's free *and* there are, increasingly, graphical front-ends for it means that -- in these times -- when you need to find places to save money, an SPSS site-license starts to look like a good bit of fat to cut. Do I think that R is ready to be used as a 'drop-in' replacement for SPSS? Absolutely not! 

 

The class that I was considering porting to R is one that is taught to Masters students who may well be, ultimately, working overseas for NGOs. So these are (relatively :-) ) motivated students going to positions where the free software could be of real benefit because the updates to libraries and to the core are all small and incremental. The last time I installed SPSS on my laptop it was well over 300MB and then proceeded to demand multiple updates to Java to boot! 

 

Regardless I haven't done the migration yet, and I may never do it. The rise of R is simply a trend that I've noted in the research and teaching that I've seen recently which got me thinking about the opportunity.

Also SPSS will get students a job afterwards: I’m not so sure about R.  

 

I agree 100% when you're dealing with a standard HR-type who has a checklist of requirements for a graduate applicant and doesn't understand that there's more than one way to skin the stats cat. They will look and say "Oh well, no SPSS." But for my own work, for instance, I actually *need* a PhD student who can use R because SPSS doesn't appear to do the mix of aspatial and spatial stats that my research area tends to require. Yes, that's a special requirement, but I'm hardly doing anything all that 'out there' and it makes a huge difference to replicability when we can run/re-run the entire analysis in one application.

 

Anyway, I'd certainly like to emphasise that I don't disagree with you, but am simply noting that R is making enormous inroads across a range of disciplines. The fact that it's free and open source makes it an interesting addition to the state-of-play in academia, regardless of whether we opt to stick with SPSS or migrate.

 

Warm regards!

 

jon

--

Jon Reades

 

Room K7.37, Strand Campus

Department of Geography
email: [hidden email]
mobile: [phone]

landline: 0207.848.1372

 

[Aidan Kelly]

Following on from John Halls comments, software and stats are one thing - getting students to accept the power, relevance and importance  of the quantitative analysis of the social is another. Can we have more contributions on  the latter problem?

e.g. when looking at the relationship between education and social class controlling for gender education has a stronger association with social class amongst is females. What are the other routes to high status occupations that men enjoy? Is there a data set that will allow us to explore this?

I think it was Gordon Marshall that said that typically social scientists only explain about 25% of the variation/variance. Anyone got some examples of where its better than this? Where it's powerful enough to convince students its really worth doing quant because that's where the good explanations can be found.

The quant initiative needs to develop a new and improved quantitative social science: technical ability in stats/software is just the means to this end.  We need better ways to conceptualise the quantitative social, more thoughtful sophisticated measures and more complex theoretical notions of causality. The software and stats is relatively easy compared to the thinking required.

Aidan Kelly BA MA
Senior Lecturer in Social Research Methods
Department of Sociology,
Goldsmiths, University of London
Lewisham Way
London SE14 6NW


Email: [hidden email]
Office: WT1106
Landline: 02070785019
Website: http://www.gold.ac.uk/sociology/staff/kelly/
Latest Publication:
http://www.tandfonline.com/doi/abs/10.1080/17449359.2013.837829#.UpXNICcqvg0

---

From: [hidden email] <[hidden email]> on behalf of John F Hall <[hidden email]>
Sent: 22 January 2014 19:38
To: 'Nicole Janz'; [hidden email]
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Nicole

 

This needs to be fed to the SPSS list as well (no embedded links or attachments allowed). 

 

Social science students are interested in social questions, not formulae or programming languages.  SPSS syntax is far more like English language and is far quicker and easier for basic operations.  In any case, as my old boss the late Dr Mark Abrams used to say, “If it’s worth saying, you can say it in percentages.” although I suspect he would accept some of the graphic output from SPSS as an alternative.

 

John

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of Nicole Janz
Sent: 22 January 2014 20:02
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Dear all,

 

the Cambridge Social Sciences Research Methods Centre (SSRMC) has changed most basic stats courses to R last academic year. I was the one converting the courses from SPSS to R (exercises, exams, lectures) and we did a 'trial' to see how it goes: A colleague taught the same courses in SPSS, while I taught in R. We did this with Bivariate Statistics and Regression. From our experience, which includes class interaction, student evaluations and feedback as well as exam results, there was no difference for the majority of students! We now teach all basic courses in R since students can use their own laptops in class. Students have accepted it. (More specifically, we use Rstudio using a prepared Rscript in class.) We also switched the main textbook from Andy Field's SPSS version to Field's R version.

 

I do not want to pass judgement on SPSS since I only used it for a year or so, but I'd say there is no disadvantage for a stats beginner learning R, and if they students want to go on to more flexible modelling and data handling later they have a head start.  

 

I do see though that anyone who has worked for years in another software which is click-based, a change might sound unnecessary - especially if all materials are already prepared and well documented. It took me 4-5 days to convert each 2-hour lecture from SPSS to Rscript, and to make sure it all works.

 

If anyone wants to have a look and download materials for basic statistics for social scientists, incl. exercises and datasets in R, I make that available here: www.nicolejanz.de/teaching.html.

 

I think there's room for all kinds of software - I'm just joining the discussion to show that a switch to R is not as frightening as some think (neither for instructors nor students).

 

Best,

Nicole

 

 

 

 

 

Nicole Janz, PhD Cand.

Lecturer at Social Sciences Research Methods Centre

University of Cambridge

Department of Politics and International Studies

www.nicolejanz.de | [hidden email] | 

Skype: nicole.janz

Blog: politicalsciencereplication.wordpress.com

Twitter: @PolSciReplicate

 

 

 

 

 

 

 

 

 

 

 

 

On Jan 22, 2014, at 6:11 PM, Jacqueline Carter wrote:

Thanks to all for this discussion. 

 

I'd be tempted to forward this to the IASSIST list as they will have experience with R in teaching - also ICPSR. I'll forward their responses to this list. 

 

I also know that @cogdog - Alan Levine at Leicester - has been using R with his students (not social scientists) so will ask for his comments. He and I talked previously about development of reusable resources for teaching with R. 

 

Best,

Jackie 

Sent from my iPhone

 

Dr Jackie Carter

UK Data Service: Director for Communications and Impact

University of Manchester Q-Step Co-director 

Mimas Senior Manager 

 

07747460963

@JackieCarter

On 22 Jan 2014, at 17:36, "John F Hall" <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email]  

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:tommacnova.edu]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice 954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall
Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]
Sent: 22 January 2014 12:19
To: Ahern, Samantha
Cc: John F Hall; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

 

[John F Hall]

The attached pdf (a slide-show) compares R with SPSS and is of genuine interest. Everyone raves about R, but students have neither interest in programming nor time in their crowded syllabus to learn it. SPSS, when used with survey data, kindles immediate interest and students can be up and running very quickly. There’s parallel debate on the QM teaching list, but it’s more to do with the place of quants in sociology, although it arose from the R/SPSS debate. John HallWhy_I_love_R.pdf

[Juanjo Medina]

Hi Aidan,

Technical ability in stats matter. It's, for example, what prevent us from making R square a simplistic measure of progress in the social sciences (see here: http://andrewgelman.com/2013/06/08/using-trends-in-r-squared-to-measure-progress-in-criminology/). But I do take your point. A lot of work about mainstreaming stats/data analysis is precisely aiming to do what you are discussing and a lot of the work of making data analysis courses more relevant to theoretical discussions also aim to do this.

Best

On Wed, Jan 22, 2014 at 9:45 PM, Aidan Kelly <[hidden email]> wrote:

Following on from John Halls comments, software and stats are one thing - getting students to accept the power, relevance and importance  of the quantitative analysis of the social is another. Can we have more contributions on  the latter problem?

e.g. when looking at the relationship between education and social class controlling for gender education has a stronger association with social class amongst is females. What are the other routes to high status occupations that men enjoy? Is there a data set that will allow us to explore this?

I think it was Gordon Marshall that said that typically social scientists only explain about 25% of the variation/variance. Anyone got some examples of where its better than this? Where it's powerful enough to convince students its really worth doing quant because that's where the good explanations can be found.

The quant initiative needs to develop a new and improved quantitative social science: technical ability in stats/software is just the means to this end.  We need better ways to conceptualise the quantitative social, more thoughtful sophisticated measures and more complex theoretical notions of causality. The software and stats is relatively easy compared to the thinking required.

Aidan Kelly BA MA
Senior Lecturer in Social Research Methods
Department of Sociology,
Goldsmiths, University of London
Lewisham Way
London SE14 6NW


Email: [hidden email]
Office: WT1106
Landline: 02070785019
Website: http://www.gold.ac.uk/sociology/staff/kelly/
Latest Publication:
http://www.tandfonline.com/doi/abs/10.1080/17449359.2013.837829#.UpXNICcqvg0

---

From: [hidden email] <[hidden email]> on behalf of John F Hall <[hidden email]>

Sent: 22 January 2014 19:38
To: 'Nicole Janz'; [hidden email]
Cc: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Nicole

 

This needs to be fed to the SPSS list as well (no embedded links or attachments allowed). 

 

Social science students are interested in social questions, not formulae or programming languages.  SPSS syntax is far more like English language and is far quicker and easier for basic operations.  In any case, as my old boss the late Dr Mark Abrams used to say, “If it’s worth saying, you can say it in percentages.” although I suspect he would accept some of the graphic output from SPSS as an alternative.

 

John

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of Nicole Janz

Sent: 22 January 2014 20:02

To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Dear all,

 

the Cambridge Social Sciences Research Methods Centre (SSRMC) has changed most basic stats courses to R last academic year. I was the one converting the courses from SPSS to R (exercises, exams, lectures) and we did a 'trial' to see how it goes: A colleague taught the same courses in SPSS, while I taught in R. We did this with Bivariate Statistics and Regression. From our experience, which includes class interaction, student evaluations and feedback as well as exam results, there was no difference for the majority of students! We now teach all basic courses in R since students can use their own laptops in class. Students have accepted it. (More specifically, we use Rstudio using a prepared Rscript in class.) We also switched the main textbook from Andy Field's SPSS version to Field's R version.

 

I do not want to pass judgement on SPSS since I only used it for a year or so, but I'd say there is no disadvantage for a stats beginner learning R, and if they students want to go on to more flexible modelling and data handling later they have a head start.  

 

I do see though that anyone who has worked for years in another software which is click-based, a change might sound unnecessary - especially if all materials are already prepared and well documented. It took me 4-5 days to convert each 2-hour lecture from SPSS to Rscript, and to make sure it all works.

 

If anyone wants to have a look and download materials for basic statistics for social scientists, incl. exercises and datasets in R, I make that available here: www.nicolejanz.de/teaching.html.

 

I think there's room for all kinds of software - I'm just joining the discussion to show that a switch to R is not as frightening as some think (neither for instructors nor students).

 

Best,

Nicole

 

 

 

 

 

Nicole Janz, PhD Cand.

Lecturer at Social Sciences Research Methods Centre

University of Cambridge

Department of Politics and International Studies

www.nicolejanz.de | [hidden email] | 

Skype: nicole.janz

Blog: politicalsciencereplication.wordpress.com

Twitter: @PolSciReplicate

 

 

 

 

 

 

 

 

 

 

 

 

On Jan 22, 2014, at 6:11 PM, Jacqueline Carter wrote:

Thanks to all for this discussion. 

 

I'd be tempted to forward this to the IASSIST list as they will have experience with R in teaching - also ICPSR. I'll forward their responses to this list. 

 

I also know that @cogdog - Alan Levine at Leicester - has been using R with his students (not social scientists) so will ask for his comments. He and I talked previously about development of reusable resources for teaching with R. 

 

Best,

Jackie 

Sent from my iPhone

 

Dr Jackie Carter

UK Data Service: Director for Communications and Impact

University of Manchester Q-Step Co-director 

Mimas Senior Manager 

 

07747460963

@JackieCarter

On 22 Jan 2014, at 17:36, "John F Hall" <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email]  

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:[hidden email]]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice <a href="tel:[phone]" value="[phone]" target="_blank">954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall

Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]

Sent: 22 January 2014 12:19
To: Ahern, Samantha

Cc: John F Hall; [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

 

--

Juanjo Medina, PhD
Senior Lecturer in Criminology
University of Manchester
http://www.manchester.ac.uk/research/juanjo.medina/

Blog: De delitos y penas

Twiter: @Juan_JoseMedina

[Aidan Kelly]

Dear Juanjo,

R square is a PRE stat - proportionate reduction in error - it uses the idea of prediction of assess explanatory power. This is the basis of most of the good measures of association excepting those based on chi square. If you reject R Square you are also rejecting these other stats too.

If we are not using explanatory power to evaluate our theory-informed models what criteria are we using to evaluate them. Not statistical significance surely!
If we are developing models of fear of crime surely the better models are, given they are theoretically satisfactory, those that explain more of the variance.
There are problems e.g. Blalock's note that the best prediction of a dependent variable is the same variable measured at a previous point in time. A level grades and degree class for example.

There are variables that are good predictors that have little or no theoretical basis - car ownership in modelling health care resource allocation is one such variable.

From a quick scan the article you cite is dealing with the practice of averaging R squares across a range of studies - I was not talking about that practice and I agree it would be an unwise thing to do.

Aidan Kelly BA MA
Senior Lecturer in Social Research Methods
Department of Sociology,
Goldsmiths, University of London
Lewisham Way
London SE14 6NW


Email: [hidden email]
Office: WT1106
Landline: 02070785019
Website: http://www.gold.ac.uk/sociology/staff/kelly/
Latest Publication:
http://www.tandfonline.com/doi/abs/10.1080/17449359.2013.837829#.UpXNICcqvg0

---

From: Juanjo Medina <[hidden email]>
Sent: 23 January 2014 09:56
To: Aidan Kelly
Cc: John F Hall; Nicole Janz; [hidden email]; [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Hi Aidan,

Technical ability in stats matter. It's, for example, what prevent us from making R square a simplistic measure of progress in the social sciences (see here: http://andrewgelman.com/2013/06/08/using-trends-in-r-squared-to-measure-progress-in-criminology/). But I do take your point. A lot of work about mainstreaming stats/data analysis is precisely aiming to do what you are discussing and a lot of the work of making data analysis courses more relevant to theoretical discussions also aim to do this.

Best

On Wed, Jan 22, 2014 at 9:45 PM, Aidan Kelly <[hidden email]> wrote:

Following on from John Halls comments, software and stats are one thing - getting students to accept the power, relevance and importance  of the quantitative analysis of the social is another. Can we have more contributions on  the latter problem?

e.g. when looking at the relationship between education and social class controlling for gender education has a stronger association with social class amongst is females. What are the other routes to high status occupations that men enjoy? Is there a data set that will allow us to explore this?

I think it was Gordon Marshall that said that typically social scientists only explain about 25% of the variation/variance. Anyone got some examples of where its better than this? Where it's powerful enough to convince students its really worth doing quant because that's where the good explanations can be found.

The quant initiative needs to develop a new and improved quantitative social science: technical ability in stats/software is just the means to this end.  We need better ways to conceptualise the quantitative social, more thoughtful sophisticated measures and more complex theoretical notions of causality. The software and stats is relatively easy compared to the thinking required.

Aidan Kelly BA MA
Senior Lecturer in Social Research Methods
Department of Sociology,
Goldsmiths, University of London
Lewisham Way
London SE14 6NW


Email: [hidden email]
Office: WT1106
Landline: 02070785019
Website: http://www.gold.ac.uk/sociology/staff/kelly/
Latest Publication:
http://www.tandfonline.com/doi/abs/10.1080/17449359.2013.837829#.UpXNICcqvg0

---

From: [hidden email] <[hidden email]> on behalf of John F Hall <[hidden email]>

Sent: 22 January 2014 19:38
To: 'Nicole Janz'; [hidden email]
Cc: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Nicole

 

This needs to be fed to the SPSS list as well (no embedded links or attachments allowed). 

 

Social science students are interested in social questions, not formulae or programming languages.  SPSS syntax is far more like English language and is far quicker and easier for basic operations.  In any case, as my old boss the late Dr Mark Abrams used to say, “If it’s worth saying, you can say it in percentages.” although I suspect he would accept some of the graphic output from SPSS as an alternative.

 

John

 

From: [hidden email] [mailto:[hidden email]] On Behalf Of Nicole Janz

Sent: 22 January 2014 20:02

To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Dear all,

 

the Cambridge Social Sciences Research Methods Centre (SSRMC) has changed most basic stats courses to R last academic year. I was the one converting the courses from SPSS to R (exercises, exams, lectures) and we did a 'trial' to see how it goes: A colleague taught the same courses in SPSS, while I taught in R. We did this with Bivariate Statistics and Regression. From our experience, which includes class interaction, student evaluations and feedback as well as exam results, there was no difference for the majority of students! We now teach all basic courses in R since students can use their own laptops in class. Students have accepted it. (More specifically, we use Rstudio using a prepared Rscript in class.) We also switched the main textbook from Andy Field's SPSS version to Field's R version.

 

I do not want to pass judgement on SPSS since I only used it for a year or so, but I'd say there is no disadvantage for a stats beginner learning R, and if they students want to go on to more flexible modelling and data handling later they have a head start.  

 

I do see though that anyone who has worked for years in another software which is click-based, a change might sound unnecessary - especially if all materials are already prepared and well documented. It took me 4-5 days to convert each 2-hour lecture from SPSS to Rscript, and to make sure it all works.

 

If anyone wants to have a look and download materials for basic statistics for social scientists, incl. exercises and datasets in R, I make that available here: www.nicolejanz.de/teaching.html.

 

I think there's room for all kinds of software - I'm just joining the discussion to show that a switch to R is not as frightening as some think (neither for instructors nor students).

 

Best,

Nicole

 

 

 

 

 

Nicole Janz, PhD Cand.

Lecturer at Social Sciences Research Methods Centre

University of Cambridge

Department of Politics and International Studies

www.nicolejanz.de | [hidden email] | 

Skype: nicole.janz

Blog: politicalsciencereplication.wordpress.com

Twitter: @PolSciReplicate

 

 

 

 

 

 

 

 

 

 

 

 

On Jan 22, 2014, at 6:11 PM, Jacqueline Carter wrote:

Thanks to all for this discussion. 

 

I'd be tempted to forward this to the IASSIST list as they will have experience with R in teaching - also ICPSR. I'll forward their responses to this list. 

 

I also know that @cogdog - Alan Levine at Leicester - has been using R with his students (not social scientists) so will ask for his comments. He and I talked previously about development of reusable resources for teaching with R. 

 

Best,

Jackie 

Sent from my iPhone

 

Dr Jackie Carter

UK Data Service: Director for Communications and Impact

University of Manchester Q-Step Co-director 

Mimas Senior Manager 

 

07747460963

@JackieCarter

On 22 Jan 2014, at 17:36, "John F Hall" <[hidden email]> wrote:

Tom

 

Thanks for joining in this ongoing debate.  It started on the QM teaching list, to which I’m also sending this reply (direct as copies have to be resent by NCRM at Southampton).  People need to know about your SPSS video as well.  (The link from my site no longer connects to your “Video tour through, and commentary on, SPSS” - with mobile screenshots)  Is it still there?  If so, do you have a new URL for it?

 

I doubt if there’s anything for R which approaches the on-line SPSS resources listed on my page (http://surveyresearch.weebly.com/on-line-spss-intros-and-tutorials.html) for social science students with little or no interest in, or previous experience (let alone pathological fear)  of maths and statistics.

 

SPSS is used by millions worldwide: how many use R?  Also SPSS will get students a job afterwards: I’m not so sure about R.  Many academic institutions are switching to Stata to save money, little thinking about the time and cost to teachers for re-tooling.  I’m sticking with SPSS and glad I’m not teaching any more, except via my website where everything is still free.

 

John

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email]  

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

 

 

 

 

From: Dr. Thomas W. MacFarland [mailto:[hidden email]]
Sent: 22 January 2014 15:01
To: John F Hall; [hidden email]
Cc: Dr. Thomas W. MacFarland
Subject: RE: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Everyone:

 

On the topic of SPSS v R, well of course each program has great value.

 

As I've posted to this list before, I first used SPSS in the mid-1980s and similar to Mr. Hall I would not want to rewrite the many existing SPSS routines that I use for this office (Institutional Effectiveness) to R or any other language

 

However, for any new project where there is no existing set of syntax, I now generally use R -- just in case the licensing fee becomes too much for those outside of this office with decision-making authority and we need to make quick transitions. 

 

Saying this, I've posted below a set of URLs for R-based online videos from a course that I currently teach on biostatistics.  The general theme is to start with the R GUI (R Commander), transition to RStudio as a middle approach to using syntax, and eventually move to R syntax only.  To use the current phrase -- git r done.

 

Comment:  The videos are open to anyone and do not require a password.  The videos are course specific so you can zip through parts of no direct interest by using the slider at the bottom of the screen.

 

Best wishes.

 

Tom MacFarland

 

Module 00 -- Video Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3i 1:14:26 Hrs

                            

Module 01 -- Overview of Biostatistics (DEP 5001), https://tegr.it/y/1bv3z 1:26:50 Hrs

                                                        

Module 02 -- Introduction:  Biostatistics and R, https://tegr.it/y/1bvmt 0:57:37 Hrs

                            

Module 03 -- Data in the Large, https://tegr.it/y/1bvmq 0:47:31 Hrs

                            

Module 04 -- Population, Normal Distribution, and Sampling, https://tegr.it/y/1bvmn 0:52:14 Hrs and https://tegr.it/y/1c1tj 1:02:24 Hrs

                             

Module 05 -- R Graphical User Interface (R Commander, R-GUI) and R Command Line Interface (R Syntax, R-CLI), https://tegr.it/y/1bvmk 0:46:54 Hrs

                            

Module 06 -- Data Exploration, Descriptive Statistics, and Measures of Central Tendency, https://tegr.it/y/1bvmh 0:55:37 Hrs

                            

Module 07 -- Student's t-Test for Independent Samples, https://tegr.it/y/1bvme 1:09:43 Hrs

                            

Module 08 -- Student's t-Test for Matched Pairs, https://tegr.it/y/1bvmb 0:36:05 Hrs

                            

Module 09 -- Analyses and Graphics for a Large Dataset With No Missing Data, https://tegr.it/y/1bvm8 0:19:36 Hrs

                            

Module 10 -- Analyses and Graphics for a Large Dataset With Missing Data, https://tegr.it/y/1bvm5 0:45:51 Hrs

                            

Module 11 -- Oneway Analysis of Variance (ANOVA), https://tegr.it/y/1bvm2 1:00:34 Hrs

                            

Module 12 -- Twoway Analysis of Variance (ANOVA), https://tegr.it/y/1btlg 1:05:55 Hrs

                            

Module 13 -- Correlation and Linear Regression, https://tegr.it/y/1bvly 0:56:30 Hrs

                            

Module 14 -- Future Actions and Next Steps, https://tegr.it/y/1bvlv 0:33:38 Hrs

 

Not a shameless plug, but I should mention that the videos are based on the use of:

 

MacFarland, Thomas W. (2014). Introduction to Data Analysis and Graphical Presentation in Biostatistics with R. New York: Springer. ISBN 978-3-319-02531-5

 

<a href="http://www.springer.com/statistics/life&#43;sciences,&#43;medicine&#43;%26&#43;health/book/978-3-319-02531-5" target="_blank">http://www.springer.com/statistics/life+sciences,+medicine+%26+health/book/978-3-319-02531-5

 

 

 

 

 

 

 

 

----------

Thomas W. MacFarland, Ed.D.

Senior Research Associate; Institutional Effectiveness and Associate Professor

Nova Southeastern University

Voice <a href="tel:[phone]" value="&#43;19542625395" target="_blank">954-262-5395 [hidden email]

 

From: SPSSX(r) Discussion [[hidden email]] On Behalf Of John F Hall

Sent: Wednesday, January 22, 2014 7:11 AM
To: [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Samantha

 

Stick to SPSS and get OU to put pressure on IBM to reduce their licence fees.  Gradpack is better than the undergrad bundle from Pearson (which doesn’t include syntax) but you should be covered by the OU licence.

 

I had a look at R, Deducer and Commander some time ago: I cannot afford the investment in time to a: learn them and b: rewrite all my tutorials.  Same goes for Stata.  Admittedly R has fantastic graphics, but SPSS is easier to use and produces publishable output.  It also has plug-ins for R (which I’ve never used) and Python (which I have, but the code was written by Jon Peck at IBM/SPSS).  If you follow the tutorials on my site, you’ll learn a lot about process and logic as well as data manipulation and analysis.  You don’t even need access to SPSS to follow them, but it helps if you want to do the exercises.

 

I’m retired and haven’t taught since 1992, but I do get positive feedback from users of my site.  Oh, and I get a free licence as a registered academic author with IBM/SPSS.

 

Everything on my site is free.

 

John F Hall (Mr)

[Retired academic survey researcher]

 

Email:   [hidden email] 

Website: www.surveyresearch.weebly.com

SPSS start page:  www.surveyresearch.weebly.com/1-survey-analysis-workshop

 

PS  Stuff copied to quant list doesn’t get circulated unless they resend from Southampton, so I’ve sent this out direct to them.  Stuff to SPSS list can’t have attachments or embedded links, so I’ve edited the bits below.

 

 

 

From: Juanjo Medina [[hidden email]]

Sent: 22 January 2014 12:19
To: Ahern, Samantha

Cc: John F Hall; [hidden email]

Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

I'm sure they send you the licence for the current year and for as long as you are a student. But if you go freelance you will have to pay over a 1000 pounds for the licence.

 

R does not require you to learn the programming language. That is simply a myth. R Commander is a graphical user interface that was designed by John Fox in Canada to teach undergraduate social science students. Deducer is not a bad GUI either for an intro course. They work in much the same way that SPSS does. In fact, R Commander does make certain things much easier. 

 

It is true the numerical outputs you get are less formatted than with SPSS (I assume that's what you mean when talking about visual representation). But I don't think that is necessary an issue.

 

On Wed, Jan 22, 2014 at 10:59 AM, Ahern, Samantha <[hidden email]> wrote:

I'm studying an OU stats course that uses SPSS and they did send me a copy with a license for this academic year.

R is good but is problematic, it requires you to learn the programming language and unless you ask things to presented visually there is no visual representation and for a lot of non-maths and computing people this is too scary - not everyone is aiming to become a statistician they just need stats to support their main work.

---

From: [hidden email] <[hidden email]> on behalf of Juanjo Medina <[hidden email]>
Sent: 22 January 2014 10:50
To: John F Hall
Cc: [hidden email]
Subject: Re: [Quantitative Methods Teaching] {Disarmed} Stats and SPSS at OU

 

Quite surprised that OU does SPSS, still. Not very "Open" of them. If you look at all the MOOCS from the other side of the Atlantic they are all embracing R, which makes sense considering is free and therefore avoids licensing issues. How much longer are we going to act as unpaid instructors and marketing agencies for IBM?

 

I appreciate you may not want to teach your intro students coding, but R Commander does pretty much everything an intro course does... at zero cost (short of adapting the teaching materials) and an introduction to the R environment seems to me a better route for the minority that will become professional researchers.

 

 

 

On Mon, Jan 20, 2014 at 3:59 PM, John F Hall <[hidden email]> wrote:

Just came across this in the programme for ASSESS 2010 at York.  Is it on your radars?

Developing Faculty wide teaching of statistics and SPSS using web-based interactive resources by Volker Patent, Open University,

See http://www.spssusers.co.uk/Events/2010/PATENT2010.pdf

 

 

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

_______________________________________________

NCRM Quantitative Methods Teaching mailing list
[hidden email]

Email: [hidden email] to unsubscribe or [hidden email] to subscribe

 

--

Juanjo Medina, PhD
Senior Lecturer in Criminology
University of Manchester
http://www.manchester.ac.uk/research/juanjo.medina/

Blog: De delitos y penas

Twiter: @Juan_JoseMedina