Wisconsin card sorting test

Thu Apr 14 20:09:36 UTC 2011

I have gotten several direct messages about this project, so perhaps 
I should post a public update.  First, thank you to everyone who has 
expressed an interest in this.  Sadly, by the time I got the help 
that I requested and found the needed materials, my window of 
opportunity passed and I had to move on to other matters while 
leaving this project undone.  I hope to return to it at a later time, 
but I cannot say when that would be.

Aside from taking time out to write the code (greatly simplified 
since the current canonical version uses a fixed card sequence, 
unlike the original research versions), one particular roadblock I 
face would be making the stimulus materials.  Several folks have 
kindly found and pointed out the WCST materials available through 
http://pebl.sourceforge.net/battery.html ; I wish to make clear, 
however, that those materials significantly depart from the standard 
cards and are entirely unsuited to my project.  (I would call their 
program "WCST-like", not a true WCST.  I aim for something higher in 
my program.)

For those who want a fully automated WCST and can spend some money, 
please note that you may purchase a commercial (and non-E-Prime) 
version through Psychological Assessment Resources, Inc. 
(http://www4.parinc.com/ ).

Once again, thanks to everyone for your interest, and I will keep 
this thread updated whenever I should make progress.

-- David McFarlane, Professional Faultfinder

On Sep 3 2010, 12:28 pm, David McFarlane <mcfar... at msu.edu> wrote:
 > After casting about far & wide, I found a complete set of WCST
 > materials right here at my own institution.  So I am all set there
 > now, my thanks to everyone who replied.
 >
 > Now I need to finish this project and submit it to STEP.  I will
 > probably do this in stages.  As stated earlier, programming the basic
 > control logic for merely administering the test is the easy part, and
 > I have that worked out already with a mere text-only
 > keyboard-response model.  Adding the graphics and improving the
 > response methods will take a little more tedious work.  I might
 > submit a first draft once I get that far.
 >
 > Then the hard work comes when I add computation of all the customary
 > raw WCST scores.  With that accomplished I would consider the project
 > finished -- I have no hope of going the extra yard to have it convert
 > raw summary scores into demographically corrected normed scores, for
 > that users will need to consult the Wisconsion Card Sorting Test
 > Manual, Revised and Expanded, by Robert K. Heaton et al. (1993),
 > Psychological Assessment Resources, Inc., Lutz, Florida, USA, or use
 > WCST scoring software.
 >
 > -- David McFarlane, Professional Faultfinder
 >
 > At 8/31/2010 05:31 PM Tuesday, David McFarlane wrote:
 >
 > >Oh, forgot to say, if interested then please contact me directly by
 > >e-mail (off the list).
 >
 > >Thanks,
 > >David McFarlane, Professional Faultfinder
 >
 > >>For those of you who might not read beyond the first line:  I will
 > >>write a professional quality WCST in EP for no charge, if I can
 > >>only get a bit of help.  Read on if interested...
 >
 > >>Well, despite the reservations I expressed earlier, on my own time
 > >>I went ahead and threw together a working skeleton in EP for the
 > >>WCST, just to flex my E-Prime muscles.  The basic control logic is
 > >>really quite simple, and with a couple of tricks (e.g., nested
 > >>attribute references) the E-Prime design is rather sleek.
 >
 > >>I would like to complete this project to my usual professional
 > >>standards and then submit it for inclusion in the public STEP
 > >>library so that everyone can use it.  In short, I aim to make this
 > >>no less than a completely faithful computerized replica of the
 > >>standardized test, in EP.  But to do that I will need many more
 > >>specific details on the mechanics of the standardized test.  E.g.,
 > >>I have been told that the standardized test presents the "response"
 > >>cards in the same pre-ordained sequence for each run, and I need to
 > >>know that sequence.  Also, if I can get hold of the scoring rules
 > >>then I might even have a whack at that, although I think it enough
 > >>just to have the program administer the test & record raw
 > >>data.  (It would also be nice if I could get someone else to do the
 > >>card graphics, as I am lousy at graphics -- I would need only 16
 > >>image files, not all 64, since I know how to make EP change the
 > >>colors.)  I should be able to learn all I need from the official
 > >>WCST Manual, with perhaps a look at a printed card deck.  But a
 > >>manual costs $115, and a card deck $170, more than I care to spend
 > >>for something that is, for me at least, only a hobby project.
 >
 > >>So here's the deal:  If any of you can help me get access to
 > >>materials that fully specify the mechanics of the test (ideally the
 > >>printed Manual and/or cards) then I will complete this project in
 > >>short order and you will all have a professional quality WCST in EP.
 >
 > >>-- David McFarlane, Professional Faultfinder
 >
 > >>At 8/23/2010 05:24 PM Monday, David McFarlane wrote:
 > >>>Well, I toyed with the idea of making a simple WCST in E-Prime
 > >>>just to amuse myself.  But when I looked into it further, I got puzzled.
 >
 > >>>The WCST started off with simple printed materials administered
 > >>>manually by a human examiner.  Clearly the WCST does not require
 > >>>millisecond precision.  So if we simply want to automate it, why
 > >>>use such an expensive, specialized, and heavyweight platform as
 > >>>E-Prime?  Wouldn't it make more sense to use some more common
 > >>>platform such as JavaScript, or Flash, or Python, or even straight
 > >>>Visual Basic?  Note that the WCST was automated using simple Turbo
 > >>>Basic (for DOS?) as far back as 1996.  Isn't this another case of,
 > >>>"When all you have is a hammer, every problem looks like a
 > >>>nail"  Or, insofar as I have not kept up with the literature on
 > >>>the WCST, am I just missing something that is obvious to the rest of you?
 >
 > >>>Also, apparently scoring the test is quite complex (perseverative
 > >>>errors, nonperseverative errors, etc.), so building that into the
 > >>>program (as opposed to leaving that to later data analysis) would
 > >>>take some care.
 >
 > >>>Finally, did anyone else know that the term "WisconsinCard
 > >>>Sorting Test" was trademarked by Wells Printing and Digital
 > >>>Services of Madison,Wisconsin, USA (see
 > >>>http://en.wikipedia.org/wiki/Wisconsin_card_sort)?  So we cannot
 > >>>give the name "WisconsinCard Sorting Test" to any printed
 > >>>materials that we produce, but since the trademark does not cover
 > >>>computerized versions we may continue to use the name "Wisconsin
 > >>>Card Sorting Test" for our computerized versions.
 >
 > >>>With all that said, note that someone did make an automated WCST
 > >>>demo for Inquisit's Millisecond
 > >>>(http://www.millisecond.com/download/samples/v3/CardSort), though
 > >>>I do not know what data it stores or how it handles the test scoring.
 >
 > >>>-- David McFarlane, Professional Faultfinder
 >
 > >>>At 8/20/2010 12:10 PM Friday, David McFarlane wrote:
 > >>>>As far as I can tell no one has made an E-Prime WCST available on
 > >>>>the Web.  It would be quite interesting to make one.  As I
 > >>>>recall, in essence the task involves operantly rewarding the
 > >>>>subject for correctly following an undisclosed rule, changing the
 > >>>>rule whenever the subject achieves an overall success criterion,
 > >>>>and seeing how well the subject can adapt to the changing
 > >>>>rules.  This would require some interesting code in E-Prime, in
 > >>>>particular scoring the success rate and then changing the
 > >>>>"correct" rule on the fly, but it could be done.  Wish I had the
 > >>>>liberty to do it myself.
 >
 > >>>>-- David McFarlane, Professional Faultfinder

-- 
You received this message because you are subscribed to the Google Groups "E-Prime" group.
To post to this group, send email to e-prime at googlegroups.com.
To unsubscribe from this group, send email to e-prime+unsubscribe at googlegroups.com.
For more options, visit this group at http://groups.google.com/group/e-prime?hl=en.