Behavior
Modification Program Steps
Deciding to Change Behavior
Beginning the Program
Evaluating the Program
Assessment
A. Describe the Target Behavior
Look for patterns in the
behavior by finding the answer to these questions. Do NOT rely on subjective
opinion but try to watch the actual behavior. An interview of the subject or of
persons familiar with the subjects' behavior may be helpful if you use focused
questions and get specific answers. The global question you are trying to answer is: under what circumstances does the behavior occur and when does it not occur? What is the pattern that the behavior displays?
- Where
does the behavior occur (only at home? at school? in the presence of
particular persons or objects?)?
- When
does it occur (time of day? day of week? weekends vs weekdays?)?
- When the
behavior occurs, how long does it endure?
- How
intense is the behavior (e.g., is the child talking or screaming)?
- How
frequently does it happen? per hour, per day, per week, per year (select
the single most meaningful period of time).
- What was
present or occurring 5-10 minutes prior to the behavior?
- What was
present or occurring within 2-3 minutes after the behavior?
- Who was
present during an instance of the behavior? Describe how these people are
related to the subject.
- Describe
in very specific behavioral terms what an instance of the behavior looks
like. Describe it so an actor could display the exact behavior. Relate
what was said as well as what was done and with what. Even seemingly
insignificant actions could provide a clue for moderating the behavior.
- Not
"He disrupted the class."
- Not
"He yelled a lot."
- Better:
"He spoke in a volume that drowned out my voice and said, "I
want my pencil back." Then he jumped on his chair and wiggled his
hips while pointing at Tim, in the right hand seat. He was laughing and
smiling throughout this period.
- Sometimes,
comments by significant others can be helpful. Favor behavioral examples
over summative evaluations.
- Not
"He annoys others."
- Better:
"He interrupts everyone who speaks within a few minutes."
- To
evaluate the specificity/observability of your description visit:
- Identifying
Behavior (obsbeh.htm).
B. Avoid these mistakes when describing the behavior
·
Confusing
motivation and behavior.
- Motives
are theories, not actual behavior. They are assumptions about the cause
of behavior. The word "because" signals an assumption. Don't
say he/she: "wanted to" "interested in"
"liked" or "disliked".
- Inferring
covert behavior from directly observable (overt) behavior. Say, "She
stared at the ceiling," NOT "she was daydreaming."
- Infusing
theories of behavior into your report. Do not use concepts like
"personality," or "insight" or "traits". It
is even an inference to say that a particular behavior is a
"habit" or "his nature." Trait names, such as
"shy," "anxious," "aggressive," also fail to
describe the behavior because they refer to theories about the cause of
the behavior.
- Evaluating
the desirability of the behavior. Behavior is of neutral value. It just
is. Its effects may be desirable or undesirable to a particular objective
or to a particular person. But when describing it, avoid those judgments.
To label a behavior as a problem or poor or good is
- vague and meaningless and thus
does not contribute to understanding it.
- imposes a set of values that
might be inappropriate upon closer inspection.
- Using
general terms such as: frequently, several, occasionally, periodically.
Instead give a precise number (e.g., 2 times per day).
C. Measure the behavior to get precise data for the above
questions.
There are a number of
methods. Several may work for any particular behavior. There is no rule to say
which is best other than that you use the method that targets the behavior you
intend to change. If, for example, you want to smoke fewer cigarettes in order
to save money on buying them, then you need to count the number of total
cigarettes consumed per day. Counting the length of time it takes you to smoke
one (while still smoking the same number) won't give you the information you
need to achieve your goal.Typically, multiple methods are used in a single project because multiple facets are of interest.
Methods:
- Frequency:
# of times a response is performed per a unit of time (e.g., day). For
example, smokes 24 cigarettes every day.
- Amount
of time: the length of time a response lasts. This may be measured in
terms of
- the
duration of the behavior (from start to finish), for example., takes
2 minutes or 20 minutes to smoke a single cigarette.
- the
number of intervals in which the behavior is observed occurring.
(e.g., only smokes during the 9am break, never at any other time = 1
interval; vs. smokes a cigarette every hour = 18 hours.) If the behavior
occurs frequently and has a clear beginning and end, then use short
intervals (10-15 seconds). Less than 5 seconds is too short-- can't tell
which interval the behavior occurred in. Don't use interval recording if
the behavior is "continuous," that is, it persists for long
periods of time with no clear beginning or end point, e.g., thumb
sucking.
- Intensity:
the magnitude or size of the response. Does the subject smoke the
cigarette down to the filter or takes a few puffs and puts it out?
- Latency:
the time that elapses until a response is performed. For example, how long
can a person goes until they pull out a cigarette.
Before beginning your assessment
You will need to answer the
following questions. - How many
times will you collect data?
- How long
will each observation period be? (long enough to get several samples of
behavior).
- When
during the day and what days of the week will you observe the behavior? Do
you need to consider weekends in addition to weekdays?
- Who and
how many observers will there be? How will you train the observer to be
accurate?
D. Identify a baseline (aka operant level) for the behavior.
Behavior modification is
about real change. You cannot determine if real change has occurred
unless you know what behavior is typical. Typical behavior is the baseline
against which the success of your intervention is measured. If you fail to
collect baseline data, then you have no way, let me repeat that, no way
to tell if your intervention worked. Behaviorists do not rely on memory which
is fallible. Without baseline data you do not have a behavior modification
program. Identifying a baseline means you collect data over a period of time without trying to change the behavior.
How long do you collect data? It depends on the characteristics of the behavior. In general, you collect enough that the behavior of interest shows a steady pattern. With some animal behaviors, that might be 1 hour. With some human behaviors, it may take several weeks.
Deciding to Change Behavior
In this phase, commitment
to the program is developed and the groundwork for a successful program is
laid.
A. All significant parties are involved and demonstrate
commitment
Encourage participation in decisions
by the subject and persons who will might have an impact on the success or
failure of the program including parents, teachers, administrators, spouses,
children, bosses, coworkers, etc.
B. Be specific and precise about the behavioral goals.
As a number of people are
involved, clear communication is critical. Even if it is a self-change, writing
our your goals and activities will ensure that you have actually been clear in
your planning. It can be a self-check.Clarity is a process of writing and rewriting. It is common, despite great effort to clarify, to discover that some behaviors remain unclarified once the program is begun. Nevertheless, do your best and it will pay off in a more successful outcome.
Include any conditions or restrictions.
Example of a precise goal:
- Not:
"To eat healthily"
- Better:
"To reduce the number of snacks (defined as ice cream, candy, or
Twinkies) from one with each meal and one in the evening to one every 3rd
day and to increase the number of vegetable portions (as defined by
Government standards) from one with my evening meal, to 6 portions per
day."
C. What are the ethical considerations?
Are there any dangers to
the subject or others? For example, woman in an abusive relationship may be
accused of trying to be seductive if she loses weight and be beaten as a
consequence. Current health status impacts on the safety of exercise and eating
programs. A person exercising in a solitary place needs to consider safety
factors.Are humane methods being used with animals? Punishment should be used very cautiously and only with close supervision.
D. Consider the total context of the behavior.
What is initiating and
maintaining a behavior (or the absence of a behavior)? Another person may
benefit in some way from the target behavior and undermine change efforts. For
example, a spouse may want someone to eat comfort foods with her and feel
neglected or judged if the subject starts eating healthily. (If you say,
"The spouse can go on a diet too" consider the ethical issues above.
The spouse may not wish to and that would be imposing values and the person
would probably simply undermine efforts.) Some persons are rewarded by
another's perceived difficulty. They will seek to continue getting their
reinforcement.
Beginning the Program
A. Identify potential interventions and select one or
several that match the target behavior.
Various procedures can be
used together for maximum effect, although a program that is too complex is in
danger of not being followed. Strike a balance between every possible procedure
and too few.Categorize your target behavior as one of the following, then see your textbook for appropriate interventions.
- teach a
never before performed behavior (reinforcement: positive and negative )
- increase
or strengthen an existing behavior (reinforcement, contingency
contract, token economy, modeling)
- extend
an existing behavior
- to
a new environment (stimulus generalization; stimulus control; modeling
)
- to
new behaviors (response generalization; shaping; chaining; fading;
prompting; modeling)
- over
time (maintenance; intermittent reinforcement; modeling)
- narrow
an existing behavior to limited environments (e.g., only snacking in the
kitchen) (also discrimination training; modeling)
- reduce
or eliminate the display of an existing behavior (extinction; time-out;
response cost; desensitization; reinforcement of incompatible responses;
modeling; punishment;)
- Identify
appropriate reinforcers for the individual
- Specify
the conditions under which reinforcement can be earned.
- You
already trained any needed data collectors during the baseline. Continue
to collect data throughout using the same methods.
- Apply
intervention. Persist with intervention until
- change
occurs,
- it
is clear that change is not going to occur and the method needs
evaluation and refinement
- On to
evaluation.
Evaluating the Program
Extensive planning increases
the odds that a behavior change program will result in behavior change if
appropriately implemented. We cannot know if that has succeeded until we
measure the behavior and compare it to the baseline.
A. Graph the results
Most data collection can be
graphed (occasionally a table is more appropriate). Graphs quickly reveal
progress or lack thereof. They allow for evaluation of hypotheses as to what
happened (or didn't). Small variations in behavior are normal. Judge progress
based on viewing multiple data collection periods (that might mean, for
example, looking at a week's worth of data, graphed by days).
B. Consider trying a "Reversal"
To demonstrate true control
over the behavior, remove the intervention. If the behavior returns to
baseline, then the chosen intervention and not some other event is the likely
cause of the change. (Of course, data is continuously collected). With some behaviors, reversals are not ethical (head banging in autistic children) or possible (learning to speak a language).
C. Evaluate the results and reach conclusions.
Conclusions will be similar
to one of the following:- The
intervention was successful in producing change as shown by......
- The
intervention was not successful in producing change as shown by.....
- The data
offer a mixed picture. These elements were successful as shown by.....
these elements were not as shown by....
Behavior Modification is a science. Clear communication of conclusions and possible implications is part of any science.
