Dear David,
I agree, up to a point. :) Yes, you do not have to use statistics all the
time. Statistics are a rite of passage for a good reason in those fields, I
would not call it silly (in my department, not only learning stats was a
rite of passage, but also a basic understanding of qual research methods
was part of the whole package, but again it depends) . Depends on the
questions that you ask.
Discovering faults (I think we discussed this before), is a different
animal. You just need a couple of examples to prove that something is not
working. Discovering why something works, on the other hand, is very
different. And there I think , you need stats (the types of clinical
practice that you mention, always are backed up with statistical studies
later on, one or two successful cases may be the product of pure chance) .
I am not, and will never be a quantoid orthodox (although quantitative
stuff is what I do, most of the time). Question drives the method, not the
other way around. But that said, the scale is off-balance in design (I am
specifically talking about the studio education literature), towards not
even qualitative research but anecdotal evidence. That, I cannot accept. I
think any field (even economics and healthcare) needs a healthy mixture of
qual and quantitative research methods. When I teach research methods to
graduate, I do not divide it as qual and quant. I teach basics, data
collection, and data analysis.
Back to topic, I still stand behind what I say. We need more long term,
longitudinal studies that use quantitative methods in design education
research.
All the best,
ali
On 11 May 2018 at 09:14, [log in to unmask] <
[log in to unmask]> wrote:
> Hi Ali,
>
> Thanks for your comment. I’ve been doing little studies of the type you
> describe for a long time.
>
> They don’t have to be statistical to have validity. That is only necessary
> if what you want to do is representative of an entire population and the
> differences you are looking for are quantifiable by their nature. Often we
> are interested in a qualitative difference between using one approach
> rather than another or between one type of outcome and another.
>
> Unfortunately, prevailing research teaching in universities in areas like
> education, social science, psychology, and marketing treat statistical
> methods as a rite of passage. You have to use them to prove you are a
> researcher. Silly stuff.
>
> One of the most important early papers I published about new methods in
> design and design education had no statistics .
> Sless, David. “Image Design and Modification: An Experimental Project in
> Transforming.” Information Design Journal 1, no. 2 (1979): 74–80.
>
> It’s now become part of our standard practice in discovering faults in
> designs. In that context we do count the number of faults at each iteration
> and we look to see a reduction over the repeat testing and refinement. But
> we don’t have to do a statistical analysis to know whether or not we have
> eliminated faults or not. As I’ve said before on this list. it’s a bit like
> clinical practice in medicine where you look for symptoms of pathology and
> then apply a treatment. You then look to see if the symptoms disappear.
>
> BTW, I was told anecdotally by a colleague, Clive Richards, that he used
> to use the method in his own teaching. Might be worth a try.
>
> David
> --
>
>
>
>
>
>
>
-----------------------------------------------------------------
PhD-Design mailing list <[log in to unmask]>
Discussion of PhD studies and related research in Design
Subscribe or Unsubscribe at https://www.jiscmail.ac.uk/phd-design
-----------------------------------------------------------------
|