Doing research in Behaviour Modification
-A minimal behaviour modification program has four phases:
1) Screening phase- for clarifying the problem and determining who should treat
2) Baseline phase- for determining the initial level of the behavior prior to the
3) Treatment phase- in which the intervention strategy is initiated
4) Follow up Phase- for evaluating the persistence of the desirable behavioral
changes following termination of the program.
Reversal- Replication (ABAB) design
- Trying to see whether the program itself actually allocated to the improvement
of the individual’s behaviour (Billie’s improved behaviour at solving math
- Reversal-replication research design- it includes a reversal (going back to
the baseline to see if it was the actually treatment that made an improvement
on behaviour) to baseline conditions followed by replicated of the treatment
phase (and, it is hoped, of the effect).
- The baseline is abbreviated “A” and the treatment condition “B”. Hence the
treatment is called ABAB design.
- Questions that might be encountered when conducting the ABAB design:
1) How long should the baseline treatment be?
2) How many reversals and replications are necessary?
- When trying to figure out which baseline graphs are the best and most
adequate, you look at the pattern of behaviour that appears to be stable and
predictable. Also you look to see the trend is in a direction opposite to the
effect predicted for the independent variable acting on the dependent variable.
Ideally then, a baseline phase should continue until the pattern of performance
is stable or until it shows a trend in the direction opposite to that predicted
when the independent variable is introduced.
- It is unethical to prolong the baseline phase if it’s in the context of that the
targeting behaviour is an abusive one.
- For the second question, if one observed a very large effect when the
independent variable is introduced, and if the area is one that has been
explored before, then one replication may be sufficient.
- Other combinations of factors might lead one to conduct several replications
in order to convincingly demonstrate a cause-effect relationship.
- Limitations posed on doing a reversal-replication design:
1) It may be undesirable to reverse to baseline conditions following a
treatment phase. When treating abusiveness of a child with a
developmental disability, for example, it would be ethically
unacceptable to reverse to baseline immediately following a successful