[Social-sciences-list] SPSS vs R or SAS

Fri Jul 27 12:05:20 UTC 2012

Hi all,

I've said the following in articles and on email lists so if you've 
read it before, then the delete key is easy to hand.

I agree with Katie that SPSS is (or more exactly, can be made) 
accessible. I dislike the amount of fighting required to get this 
done and feel that both R and to a lesser degree, SAS have an 
advantage on this front. R can be operated using a screen reader 
immediately after the default installation is completed.

Working in collaboration with colleagues is a key reason for fighting 
with any semi-accessible software. As a statistician, I get to choose 
what software I use and for more reasons than are relevant here, R is 
the best software for me and my work. I suggest that some software is 
worth fighting with (SAS and to a lesser degree SPSS) but have black 
flagged other options.

One key reason for choosing a particular software solution is the 
support and documentation on offer. Accessible documentation does 
exist for SAS and R with much of this being free and online. Email 
lists are useful but most novice enquiries are not appreciated and
therefore should not be relied upon. I can't say anything about the 
recent versions of SPSS and its internal help facilities. I'd 
appreciate any comments Katie and others may have on this.

One reason I do not like using SAS is the need to wade through a long 
amount of output to get to the things I need. In model building work, 
I need to look at a few bits here and there. Only after I've found 
the model I want do I then look at more of the details.

Use of menus and dialogue boxes vs use of commands is an interesting 
debate. SAS does have a well established command syntax and much of 
this code is provided once the menus are used to generate analyses. 
Other software like Minitab (and the last version of SPSS I used) do 
this too. R does not. There are front ends being developed for R, but 
they are not accessible, nor essential, as the use of commands is 
preferred by most users. A GUI is preferable for occasional users who 
do not need to reproduce their work after minor modifications. I 
believe efficiency for the blind user comes from a command based 
software option, but relies on coming to grips with the specific 
syntax. Commands in R tend to be quite concise in comparison to those 
used in SAS. Knowing what commands are relevant at the time you want 
them is a problem for all command driven options.

The LURN document mentioned in one of my previous messages was 
written for introductory statistics students, most of whom were 
Engineering students. I modified the way this was put together for 
blind users and added some extra material. It is still a work in 
progress as I do add things that people ask of me from time to time. 
See http://r-resources.massey.ac.nz for more.

I note that the suggestion of creating data for importing into SPSS 
is to use Excel. I think this is good advice for all software 
options, not just SPSS. I'm not all that excited by the idea that I'd 
have to work with Excel or Word files to read the output though. I 
know that in older versions of SPSS, users could save their output 
files as text files. These can be opened in an internet browser so 
that when the file is updated, the user can hit refresh (f5) to see 
what the most recent additions were. I have seen this approach used 
by SPSS and Stata users and have tested it thoroughly with R.

As I am not a Word user by preference, I have not investigated one 
way of working in R that I think has merit for many users. An add-on 
package (there are lots of these) can direct all output to a word 
file, including the graphs. Another option is to send the output to 
an html page which is useful for those of us in education settings 
developing web material. Directing output to a text file poses no 
problems at all.

My final word (for this message) is that there is no one way that we 
should all follow. I can demonstrate the usefulness of R over every 
other option that I've seen. I do not however, believe that R is the 
best option for everyone.

Jonathan

At 06:50 p.m. 27/07/2012, you wrote:
>Hi Maha and all,
>
>While the experience of using SPSS with JAWS is not perfectly
>semaless, I would like to point out that, with a few work-arounds,
>even the latest version of the program, SPSS 19, is quite usable for
>the blind. As a Psychology graduate student who primarily works with
>quantitative data, I use SPSS as my main analysis tool and have not
>run into any major problems. As others have already mentioned, certain
>steps are required before SPSS becomes usable with JAWS; in the case
>of version 19, it is necessary to download a patch from the IBM
>support site in addition to modifying the path variable and replacing
>the accessibility.properties file. Furthermore, it is best to import
>data from Excel into SPSS (rather then entering them directly into the
>data editor), and all output must be exported into Excel or Word
>before they can be read. I have heard that R and SAS are more easily
>accessed with JAWS, but from what I understand they both have a bit of
>a steeper learning curve since all analyses need to be performed using
>syntax (whereas most functions of SPSS can be accessed either through
>the menus or  command syntax). Another issue to consider while
>choosing a statistical package is the norms of your field of study.
>Given that most of my colleagues use SPSS, it is farily easy for me to
>find technical assistance if I'm struggling with carrying out a
>particular analysis, whereas it might be more difficult for me to find
>help if I were using a different statistical software. I have not
>worked much with qualitative data so unfortunately do not have much
>information about the accessibility of programs used to handle them.
>
>Katie
>
>
>On 7/26/12, cheryl fogle <cfogle at unm.edu> wrote:
> > Hello Maha, Faiz, and list,
> >
> > Statistics questions come across the blind science and blind math lists
> > all the time.  I'd suggest subscribing to those lists if only for access
> > to the archives.  Some questions may have already been answered.
> >
> > I used an older version of spss for qualitative and quantative analysis
> > for my dissertation.  It was version 15 because that was the one I
> > bought several years ago.  My student license expired so I'm giving
> > advice from memory.  I generally imported excel files and then selected
> > the row and column ranges for the different analysis dialog boxes.  Then
> > I always exported the output back to excel.  The results are still
> > readable and I'm still referring to them as I revise my manuscript.
> > Sorry I don't have experience with 18.
> >
> > Cheryl Fogle
> > Ph.D. candidate, Department of Anthropology, University of New Mexico
> >
> > On 7/26/2012 3:52 AM, Faiz Rasool wrote:
> >>
> >> Hi Maha,
> >>
> >> Unfortunately, SPSS is not completely accessible with jaws. It is
> >>  claimed by IBM that from version 18 you can use the software by
> >> making few changes in the path variable and replacing the
> >> accessibility. Properties file. After making  the suggestive changes
> >> to improve the accessibility of SPSS with jaws, I found that I could
> >> use the menus  e.g. open the  analyze menu to request for a cross
> >> tabulation or run a simple linear regression. But  few aspects of the
> >> programs e.g.   entering data directly in the data editor window of
> >> SPSS was not possible  at least for me. When you try to use the arrow
> >> keys to navigate in the data editor windows of SPSS jaws does not
> >> always correctly report the location of the row  and column you are
> >> in. The workaround I found was to first enter my data in excel, and
> >> then import it to SPSS. Please note that I did not experienced any
> >> major difficulty in assigning variable and value labels  in SPSS. The
> >> second major issue was that   visually impaired users cannot directly
> >> read the content of the output window of SPSs. Your option to read the
> >> output window is  to  export the  output window as a  word or pdf
> >> file. Personally I found the option of exporting as a word file very
> >> helpful as I could copy and move the tables in other files e.g. course
> >> assignments. Please note that other users have reported few problems
> >> in setting up the location of the path variable and replacing the
> >> accessibility.properties file. Dr. Jonathan who's on the list has
> >> always   maintained that R is the best software option for blinds. I
> >> personally have very limited experience of using R so I cannot offer
> >> much advice myself.   But I think Dr. Jonathan has maintained a
> >>  website where information on using R can be found. I'd  suggest that
> >> you should explore R if possible. I used stata    for a limited time
> >> and found it to  work well with jaws. My problem was that  course
> >>  instructor wanted  all the assignments in MS word and I could not
> >> work out how to get the output of R or Stata directly to mS word. With
> >> workarounds I was able to get the output of SPSS to MS word though.
> >> Please note that it is not that the output of R and stata cannot be
> >> sent to MS word it was just that I could not learn how to in the
> >> limited time I had and there was no accessibility support office to
> >> come to my help.
> >>
> >> Hope this helps,
> >>
> >> Regards,
> >>
> >> Faiz.
> >>
> >> *From:*social-sciences-list-bounces at nfbnet.org
> >> [mailto:social-sciences-list-bounces at nfbnet.org] *On Behalf Of *Maha
> >> Khochen
> >> *Sent:* Thursday, July 26, 2012 11:48 AM
> >> *To:* Blind Social Scientists List
> >> *Subject:* Re: [Social-sciences-list] introduction
> >>
> >> Hi everyone,
> >>
> >> My name is Maha. I am currently reading for my MPhil/PhD at the
> >> Institute of Education, University of London- UK. My research is
> >> looking at the inclusion of visually impaired learners in mainstream
> >> education in Lebanon, my home country. It is so interesting to see
> >> this mailing list up and running.
> >>
> >> I have been looking for some advise on accessible qualitative software
> >> package for a while. And was wondering if there is anyone on the list
> >> who can advise? I have tried the NVivo but unfortunately it appeared
> >> that it is not accessible with Jaws, the screen reader that I use to
> >> access information. Any advise with this regard is very much
> >> appreciated. Note that I have no residual vision.
> >>
> >> Another thing, I am taking training on the SPSS software very soon, it
> >> would be very much appreciated if I can have some advise based on your
> >> experience whether it is accessible and that blind researchers can use
> >> it with no problem. Alternatively, any advise for an accessible
> >> quantitative software package would be very much appreciated too.
> >>
> >> Looking forward to hearing from you.
> >>
> >> Maha
> >>
> >> *From:*Morman, Ed <mailto:EMorman at nfb.org>
> >>
> >> *Sent:*Wednesday, July 25, 2012 4:01 PM
> >>
> >> *To:*social-sciences-list at nfbnet.org
> >> <mailto:social-sciences-list at nfbnet.org>
> >>
> >> *Subject:*[Social-sciences-list] introduction
> >>
> >> Dear Listmembers---
> >>
> >> I'm pleased to see that my colleague Anna Kresmer has already joined
> >> the list.  Like Anna, I am a sighted staff member at the NFB Jernigan
> >> Institute in Baltimore.  I have a PhD in History of Science and worked
> >> for many years in the field of history of medicine as a librarian,
> >> occasional teacher, and not terribly productive researcher.  I know a
> >> wee bit about social science research methods and have an avocational
> >> (extremely avocational) interest in cognitive science.  Curiosity
> >> about how people construct their internal model of the world without
> >> sight was part of the reason I was interested in working for an
> >> organization related to blindness---especially a blind-led
> >> organization of blind people.  I've been here for four years now and
> >> I'm enthusiastic about NFB philosophy and its notion of what
> >> constitutes proper training.  With Anna Kresmer and our colleague Lou
> >> Ann Blake, I've developed something of an understanding of the history
> >> of the blind in the U.S. and more specifically the history of the NFB.
> >>
> >> I'll mainly be lurking, but might occasionally have something to say.
> >> With any luck it will be intelligent and even possibly helpful to
> >> other listmembers.
> >>
> >> Cordially,
> >>
> >> Ed Morman
> >>
> >> Edward T. Morman, MSLS, PhD
> >>
> >> Director, Jacobus tenBroek Library
> >>
> >> NATIONAL FEDERATION OF THE BLIND JERNIGAN INSTITUTE
> >>
> >> 200 East Wells Street
> >>
> >> /   at Jernigan Place/
> >>
> >> Baltimore MD 21230
> >>
> >> 410.659.9314 x2225
> >>
> >> 410.685.2340 (fax)
> >>
> >> ------------------------------------------------------------------------
> >>
> >> _______________________________________________
> >> Social-sciences-list mailing list
> >> Social-sciences-list at nfbnet.org <mailto:Social-sciences-list at nfbnet.org>
> >> http://nfbnet.org/mailman/listinfo/social-sciences-list_nfbnet.org
> >> To unsubscribe, change your list options or get your account info for
> >> Social-sciences-list:
> >> 
> http://nfbnet.org/mailman/options/social-sciences-list_nfbnet.org/mahakhechen%40hotmail.com
> >>
> >>
> >>
> >> _______________________________________________
> >> Social-sciences-list mailing list
> >> Social-sciences-list at nfbnet.org
> >> http://nfbnet.org/mailman/listinfo/social-sciences-list_nfbnet.org
> >> To unsubscribe, change your list options or get your account info for
> >> Social-sciences-list:
> >> 
> http://nfbnet.org/mailman/options/social-sciences-list_nfbnet.org/cfogle%40unm.edu
> >
> >
>
>_______________________________________________
>Social-sciences-list mailing list
>Social-sciences-list at nfbnet.org
>http://nfbnet.org/mailman/listinfo/social-sciences-list_nfbnet.org
>To unsubscribe, change your list options or get your account info 
>for Social-sciences-list:
>http://nfbnet.org/mailman/options/social-sciences-list_nfbnet.org/a.j.godfrey%40massey.ac.nz

_____
Dr A. Jonathan R. Godfrey
Lecturer in Statistics
Institute of Fundamental Sciences
Massey University
Palmerston North

Office: Science Tower B Room 3.15
Phone: +64-6-356 9099 ext 7705
Mobile: +64-29-538-9814
Home Address: 52 Linton St, Palm. Nth.
Home Phone: +64-6-353 2224 (Just think FLEABAG) 