[From Rick Marken (950623.1530)]
Bruce Abbott (950423.1535 EST) --
I'd like to see the PCT model of this (Turbo Pascal, please)...If it's a
higher-level control system, then the variable involved needs to be included
in the reinforcement model, doesn't it?
I didn't realize that reinforcement theory was based on PCT;-) What have
"higher level variables" got to do with reinforcement theory? Reinforcement
theory says (I think) that responses are selected by their consequences
(reinforcements, punishments and non-reinforcements); the position of the
cursor in the R. coli task depends on responses that are presumably selected
by their consequences. How might selection of consequences produce the
results we see?
I would have to know what is reinforcing the behavior of changing the target
position
That's what I have to know, too! Any ideas about where to look for what is
reinforcing this behavior? I've described every possible consequence of
responding in this task. There were only two possible consequences of
responding; the cursor either moves away from the target after a press or
toward it.
Let's perform this experiment:...I have arranged a contingency such that the
participant earns $1.00 per second so long as the cursor stays within 0.25
cm of the visible target. The participant knows this. You are offering
nothing in return for compliance with your suggestion. Prediction?
From a PCT perspective, it depends on what the subject wants.
Further questions: What is the reinforcer for staying on target (the
visible one)?
I believe a reinforcement theorist _should_ say that the reinforcer is the
consequence of responses that increases the probability of the responses that
produce that consequence (what a reinforcement theorist actually says is
quite another matter). So it would be the cusor movements (presumably
movements toward the target) that increase the probability of the handle
movements that produce them.
What happens if the participant adopts your suggested new target position?
The cursor will stay near my suggested target position.
What caused the change in average cursor position?
Depends on what theory you buy. PCT would say the change was caused by a
change in the subject's reference specification for the perceived position of
the cursor. Reinforcement theory would say that the change was caused by a
change in the consequences of responses.
By the way, you said in an earlier post:
there is a similar independent means of determining whether a consequence is
a reinforcer, punisher, or neutral event under given conditions. Thorndike
described it in 1898.
I think it is CRUCIAL that we know how this is done. How do we determine
whether a consequence is a reinforcer?
I'm under the impression that we determine whether a consequence is a
reinforcer, punisher or neutral event under given conditions by determining
whether the consequence leads to an increase (if the consequence is a
reinforcer) a decrease (if the consequence is a punisher) or no change (if
the consequence is a neutral event) in the ambient probability of the
response that produced the consequence.
If this is wrong, please give details (quantitative if possible) on the
correct way to determine whether a consequence is a reinforcer.
Best
Rick