Greetings LOrs!
This post is really a follow up to my own posting (LO23463), but also ties
in to LO23471 from Leo Minnigh (I was prompted to right some points of
clarification regarding LO23463 by another list member in an off-line
conversation).
The first point of clarification has to do with the difference between
negative reinforcement and punishment as I used them in LO23463. These
two terms are often confused and used interchangably, but they are not
synonomous. Negative reinforcement is behavior that is "motivated" by the
desire to escape an undesired antecedent condition or to avoid a
threatened undesired outcome - in both cases, the rate and future
likelihood of the behavior increases. To illustrate, if you have a
headache (the udesired antecedent condition) you take Tylenol to "escape"
headache. In the future, assuming that the Tylenol reduces or eliminates
the headache, you will be more likely to take Tylenol in the future when
you experience a headache. Punishment, on the other hand, results when an
undesired outcome follows a behavior. This decreases the likelihood that
the behavior will recur - at least for a while).
In my post, I seemingly interchanged negative reinforcement and punishment
freely. This may have had the misleading effect of implying that they
were synonomous. I was actually trying to communicate that they are often
used in conjuction with each other (making up what I referred to as the
negative reinforcement paradigm). I probably should have taken the time
to redraw the distinction between the two (which is what I am doing here)
and rename what I referred to as the "negative reinforcement paradigm " as
an aversive controls paradigm or some other descriptive term (covering
both negative reinforcement and punishment as means of influence) that
would not cloud the issue.
At any rate, perhaps I wasn't as clear as I should have been about the
distinction. Leo, I have been digesting much over the past number of
months. However, this is not the only reason why I don't participate as
much as I would like on the list. The precision necessary to accurately
discuss and explain behavioral principles in response to a specific area
or context (the area where I most often feel that I have something to add
to the discussion) requires a great deal of time and effort on my part. In
the posts that I originally wrote to the list some time ago about
behavioral principles, I invested much of my free energy trying to be
clear in defining behavioral principles and separating the terms
consistently and clearly. One reason why I don't contribute to the list
often (and when I do, the posts are often lengthy) is because of the rigor
with which I feel compelled to write. I generally don't have the time to
think all of the issues through and be as precise as I fe! ! el is
required, and, as a result, I refrain from entering the fray most of the
time.
The second point for which I wanted to provide some extra clarification
regards the application of behavioral consequences to groups. My comments
here were not intended to suggest that we should never arrange for
behavioral contingencies at the group level. Rather, I was trying to
caution that doing this to the exclusion of considering the individual
contingencies for the members of the group may result in a great deal of
frustration. For example, when you begin to work at a level of
aggregation larger than the individual, the results may easily be
misinterpreted. The consequences for the group tend to be delayed from
the individual's behavior. The efforts and contributions of the
individuals in the group may have been varied, resulting in some of the
top contributors feeling that their efforts were punished, while the low
performers are reinforced for behaviors that may have detracted from the
group's success rather than contributing to it (the attainment of the
reinforce! ! r is not contingent on the behavior of the individual).
Finally, the "one size fits all" reinforcer that is often selected when a
group is rewarded (done in the interest of perceived "fairness") is valued
by some in the group, but not all.
Many people have attempted to provide rewards for groups of people and had
their efforts produce far different outcomes than they expected, or, even
worse, the effort completely backfired, resulting in far less of what the
intervention was trying to inspire more of. This is why most people will
respond initially to the presentation of behavioral principles by
exclaiming that they have already tried postive reinforcement, and it
doesn't work. However, in these situations, they haven't really applied
behavioral principles (their efforts violated many of the requirements for
effective implementation). They have lost the precision necessary for
making rewards reinforcing as well.
On the other hand, if you begin by examining how individuals are
responding to your attempts to reinforce, you are in the ideal position to
evaluate the success of your attempts to reinforce. If you begin at the
level of the individual, then you will have a great deal more success in
your attempts to "reinforce" a group - you have built the bridge between
the individual behaviors and contributions and the group success. But if
you don't begin with the individual level, and you don't have a high level
of understanding with the application of contingent reinforcers, when your
attempts to reinforce don't have the expected/desired effect, you most
likely will also not have any idea about what to do differently. At's
suggestion to break the class down into small learning groups with
individual peer leaders/teachers creates an environment where the
individual is much more likely to receive support and reinforcement for
their efforts than they would be in a traditional classroom (! ! one
teacher, 25 students).
In order to avoid this confusion in my work with others, I draw a
distinction between reinforcement and rewards by making these points. You
can reward a group or an individual for their outcomes, but you can only
reinforce an individual for their behaviors. Alfie Kohn has written a
couple of books that define and consider many of the problems of rewards,
and he is absolutely correct in many of his observations regarding the
potential downfalls that may result when "this [outcome] is rewarded with
that".
Where I feel that Kohn goes off track is in his conclusion that these
observations support the conclusion that behavioral principles don't
really work. In fact, the (mis)application of rewards for outcomes in
lieu of providing contingent reinforcement for behaviors is the source of
the problem, not the rewards themselves - a point which Kohn seems to miss
completely.
Alright, now that I have that off my chest, I can sleep with a clear
conscience.
Jon Krispin
--"Jon Krispin" <jkrispin@prestolitewire.com>
Learning-org -- Hosted by Rick Karash <rkarash@karash.com> Public Dialog on Learning Organizations -- <http://www.learning-org.com>