[from Gary Cziko 950106.1535 GMT]
Rick Marken (950105.1330) asserted:
If the cause-effect model is
rejected, for example, it means (as we "extremists" have said) that all of
the data collected within the framework of this point of view (ie. all
behavioral science research) is junk -- even the kind where there is a nearly
perfect relationship between presumed cause (IV) and presumed effect (DV).
While I agree that the PCT interpretation of purposeful behavior is
superior to the IV-DV one, I don't agree with Rick that all IV-DV research
is necessarily "junk." If a strong relationship is found between an IV and
DV, then this indicates that the IV is disturbing some controlled variable.
While the conventional psychologist might stop here, the PCTer can begin
to apply The Test to attempt to find out what the IV is disturbing. In
this way, PCT can follow-up on conventional IV-DV research. And knowledge
of the IV-DV relationship alone MIGHT be useful for prediction even if the
controlled variable is unknown (but only if the IV is the only factor
disturbing the controlled variable and it disturbs it always in the same
way--as my and Rick's recently posted rubber band demo show).
The trouble is, of course, that most conventional IV-DV relationships are
derived statistically over groups of individuals and may not apply to all
or even any of the individuals in the study (as Runkel and Powers have
demonstrated and argued). But reliable IV-DV relationships found over many
individuals are just the sort of thing that a PCTer should want to know
about in order to go the extra step and find the controlled variable.
Perhaps Bruce Abbott can provide an inventory of such reliable,
individual-based IV-DV relationships discovered by psychologists that we
PCTers can follow-up on.--Gary