Dear Don
No yelling but you are wrong.
e.g. in Action Research repeatability is not an issue or not possible because community action situations are unique. At best one talks of recoverability.
Best
Birger
________________________________________
Fra: PhD-Design - This list is for discussion of PhD studies and related research in Design [[log in to unmask]] på vegne av Don Norman [[log in to unmask]]
Sendt: 9. februar 2012 17:18
Til: [log in to unmask]
Emne: Yes, there is a (single) scientific method
I am posting this as a separate contribution because i do not want to
get involved in the current debate about research. But the question
was raised as to whether or not there is a (singular) scientific
method. I explain why the answer is yes, there is but one, single
scientific method.
And before the yelling starts, the scientific method is a noverarching
philosophy. It gets executed very differently in the many disciplines
and sub disciplines, but the overarching principles still remain. And
they are very simple, having two major principles:
Principle One: reproducibility.
Principle Two: Replication, confirmation or disconfirmation.
Science progresses by repeating other people's work, either confirming
or disconfirming it, and then building upon and extending those
findings for future researchers to try to reproduce, extend, or
contradict.
The Wikipedia article on this topic is surprisingly good:
http://en.wikipedia.org/wiki/Scientific_method
---
This means that any claim, whether of the effectiveness of a
procedure, or of a newly discovered observation, or whatever, is
always written and described in a way that other people can try to
reproduce the work precisely. This means that the methods are
described precisely, as are the observations, the people who might
have been involved (their demographics and the way the observations
were performed), and then the way by which observations were
interpreted, including a statistical analysis where appropriate to
ensure that the results were not a result of random variability.
Others can then try to repeat the work. If they get similar results,
then they can extend the work. If they get different results, then a
debate ensues within conferences and publications (and internet
mailing lists) and continues until some agreement is reached. (And, to
be honest, some debates in science "reach agreement" when the
promoters of one of the sides die or retire.)
The critical thing is that work can be repeated and verified. This
allows the base of knowledge to be extended over time.
---
In the field of design, many of the basic methods and beliefs have
never been subjected to this kind of scrutiny. As a result, we do not
really know if the claims that are made about them are true. Now, some
of the issues in design do not need scientific assessment. Some of
them require knowledge that does not yet exist -- such as how to
assess aesthetic values (there are many attempts to do this, but these
are still evolving). But there is nothing inherent in design that
would prevent rigor in evaluation.
Design researchers have a very wide range of methods they employ. But
for these to be a science, all should provide for explicit repetition
and comparison of results. This is already true for a good deal of
the work.
Mathematics: Note that this description says nothing about the role of
mathematics. Many people equate the scientific method with
mathematical models. No, that is a separate topic. No math is
required. Just careful description and the ability to replicate
results.
Don
Don Norman
Nielsen Norman Group
[log in to unmask] www.jnd.org http://www.core77.com/blog/columns/
IDEO Fellow. Latest book: "Living with Complexity"
|