Hi David,
Interesting observations and thanks for sharing. First, a note: I am not a
quantoid-ortodox who believes that everything can be quantified, I think
one of the hallmarks of a good statistician is to know when to stop and
look for other ways to tackle the problem :)
That said, just out of curiosity, have you ever tried mathematical models
(such as cluster analysis, multiple correspondence analysis, some forms of
machine learning etc.) to see whether there are some hidden patterns in how
the performance deteriorates? Or how these new rules are invented? Because
not all statistical/mathematical models are used for prediction and causal
analysis, sometimes they are very good for seeing patterns beyond simple
counts and percentages.
But again, I am a total stranger to your area and I am just asking out of
curiosity.
Best wishes,
ali
-----------------------------------------------------------------
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
-----------------------------------------------------------------
|