[FROM: Dennis Delprato (921020)]

David Goldstein

I work at a residential treatment center for children ages 12-17.
It is an unlocked setting. Sometimes a resident will leave center
grounds ("go AWOL"). Recently, a resident was tragically killed
during an AWOL. The problem is how to reduce AWOLS among residents
to zero.

A behavior modification approach: (a) Consequences have to be
identified which when presented after an AWOL to a resident will
result in the frequency of AWOLS decreasing. The consequences we
have used are: deduct points (relates to money), reduce status
level (relates to degree of supervision and activities allowed),
conduct a special meeting which includes the resident, treatment
team members and administrators ( ETRE meetings) which can result
in discharge, a special program, verbal warnings and lectures.

(b) Arrange the environment before an AWOL occurs which will act to>
prevent AWOLs from happening. The "stimulus control" efforts we
have used are: personal physical restraints when staff judge a
resident to be in a state which poses a risk to self/others, verbal
statements to residents to stop or return, verbal reminders to
residents reminding them of the consequences.

It is obvious from the fact that residents are still going AWOL
that the above measures are not controlling the level of AWOLs to
the desired level of zero. The measures we are taking might be
acting to reduce the AWOL frequency but it would be hard to prove.

Brief comments from my viewpoint re. modern behavioral theory and

Solutions based on (a) are generally very ineffective. Those based
on (b) as described are but dimly justified by behavioral theory;
they are more folk solutions. To label them "stimulus control"
interventions is rather gratuitous.

The above do not recognize more multivariate developments
in behavioral therapy (see Goldiamond, Behaviorism, 1974, 2,
1-84; simplification by Delprato, J. Behav. Ther. & Exp. Psychiat.,
1981, 12, 49-55; example of elaboration by Delprato & McGlynn,
ibid., 1988, 19, 199-205). I have pretty much quit following
this literature but get the impression that it is still being

If I were working on this case, I would first examine closely the
complainants, realizing that *all* clinical cases require at
least one complainant. Can the complaining be changed? YES.
Send the identified client away. NO. There are other systemic
considerations, but assume we *do*, for the present, have to
deal with AWOLs.

1. What can we help the client to do that will increase his
acceptability to others? Leaving the grounds can be used as
a reinforcer--with proper planning, gradations, etc.--for
performance of socially acceptable behaviors, including academic

2. Why does a client want to leave? Sounds like (a) facility
is less reinforcing than outside and/or (b) facility is more
aversive than outside. What can be done to make ordinary
activities more positively reinforcing and less aversive?
Why should the client want to stay there? Of course, it
is a tough problem because so many factors operate outside
and in settings such as these to keep routine interactions
anything but "rewarding."

A constructional approach (Goldiamond's term) suggests we
concentrate on building new behavioral repertoires (ah that
lingo) rather than on eliminating behaviors. Furthermore,
we are advised to begin with what the client gives us;
there must be something there from which we can help the
development process. What does the client like to do?
Read comic books? Fine, this can be incorporated into a plan
that point to development of personally and socially acceptable
behavior patterns.

Note that there is virtually no hint of constructional
interactions in (a) and (b) above. Both are all eliminative.
The basic problem with eliminative efforts (show person
what *not* to do) is that they do not specify what *to do.*
I know not to do a, b, c, d, e, ad infinitum. So what do
you want me to do? Why are you keeping it a secret?
Furthermore being "still, quite, and docile" is not what
I need to know. If a dead person can do it, it is not
useful for actually living people. So constructional
interventions do require a bit more effort on the part
of those in charge; they have to come up with *alternatives*
that they find suitable. But, hey, this is fair. They
are the representatives of the complainants (even when the
complainant is the client).

In their simplest form, constructional therapies are based on
the infamous positive reinforcement, but where the reinforcers
are more ecological rather than contrived by way of the
equally infamous "deprivation schedules." The idea is that
the world works best (witness USSR) when there are systematic
consequences to response occurrances as opposed to "freely
avaliable reinforcers." (As in few of us go through life
with incomes that are independent of our efforts--no
workee, no payee.) Actually, control system theorists may
find Goldiamond's (1974) paper of interest. He wrote it
against the background of ethical and legal concerns that
came up in the 1970s over "behavior control." Goldiamond
suggested the best way to handle such matters was to begin
(in U.S.A.) with the Constitution. Not a bad idea.

Despite my distinct lack of enthusiasm for the postulates of
behavioral therapy, I do not find superior technologies at
present. I do find that knowledge of contemporary behavior
therapy is not widespread. The solutions implied by (a) and
(b) above hark back to the 1960s. I suppose this might show,
in part, that behavior therapists do not do a very good job
of communicating new developments in their field--apart
from the incorporation of (largely outmoded information
processing) cognitive theory.

Dennis Delprato
Dept. of Psychology
Eastern Mich. Univ.
Ypsilanti, MI 48197