Marken's Correlation Illusion

[From Dag Forssell (930918 1125)]

Having sent my post last night, it was hard to shake thougts about it.
It occured to me in the middle of the night that while you write
towards the end of your post:

             ........(and rescued psychology long ago from
it's deathly addiction to IV-DV research).

old habits die hard. You are thoroughly trained in IV-DV research and use
it below to show something that happens to be true, but has no
significance. You draw false conclusions -- unless I shall prove
mistaken, of course -- and accuse others of being mistaken. Shall we call

situation where there is NO relationship between p (or r-p) and o,

                          ^^^^^^^^^^^^^^^^^^^^^^^ ^^^^^^^^^^

even though p is controlled (p~r). I just did this with my little
Hypercard control simulator. A scatter plot of temporal variations
in (r-p) against temporal variations in o looks like this:

     > x
     > x x
  o | x x
     > x x x

Not all points are plotted but this gives a representative picture
of the shape of the whole plot. The x's represent paired values of
p and o at different times. The relationship between r-p and o is
a cloud, not a function -- ie. there is no causal relationship
between r-p and o. ...

Bill Powers just wrote to Hans Blom (930916.0845 MDT):

"It is very hard to underestimate the power of statistics as used in the
behavioral sciences."

Your mistaken application of correlation analysis above, in lieu of
consideration of functional relationships, is interpreted by this
observer as an excellent illustration of Bill's thesis.

It seems one is much better off without studying statistics.

Standing by for your counter-blast in this exchange of disturbances.

Best, Dag