People in ABA

Robert Koegel and Laura Schreibman developed PRT (Pivotal Response Training).  It targets self-initiation of the subject’s preferred activities in a natural setting..

Fred S. Keller developed PSI (Personalized System of Instruction), a self-paced teaching method.

Ogden Lindsley developed precision teaching which uses a semi-logarithmic chart also known as a Standard Celeration Chart.  The chart has a slope of 34 degrees.

Stimulus Control: Salience, Masking and Overshadowing

Stimulus Salience refers to how obvious or prominent a stimulus is in a person’s environment.  If a person has visual deficits, then visual stimulus will not have as much salience as auditory stimulus, for example.  In order to notice stimulus, and for that stimulus to have salience, a learner must possess pre-attending skills necessary for the setting.  The pre-attending skills for kindergarten, for example, include looking at the instructional materials, listening to instructions and to the teacher and sitting quietly while instruction is happening.

Masking is when the salience of a stimulus is decreased.  A competing stimulus blocks the evocative power of the stimulus, decreasing its effectiveness.  For example, a teenager may follow directions when alone with a parent, have a more difficult time when peers are present.  This example is competition of different contingencies of reinforcement which makes it more difficult for the subject to mind the discriminant stimulus.

Overshadowing is when the first stimulus has no more stimulus control.  An example is a teenager who can study in a classroom, but not in front of the a group of cheerleaders.

In order to reduce the effect of overshadowing and masking, we must apply antecedent interventions such as:  arranging the environment to reduce “noise” from unwanted stimulus, making the instructional stimuli intense and consistently reinforcing behavior in the presence of desired stimulus.

Reinforcer Assessment

The quality / power of reinforcement can be determined with a reinforcer assessment.  A variety of direct, data-based methods are used to present one or more stimuli contingent on a target response and then measuring the future effects on the rate of responding.

  • Concurrent schedule of reinforcer assessment: pits two stimuli against each other to see which will produce the larger increase in responding when presented as a consequence for responding.  The more effective reinforcer is the one that has more responses.
  • Multiple schedule reinforcer assessment: reinforcers are delivered for the same behavior, on the same schedule but at different times; an SD is present to signal which schedule is in effect.  The more effective reinforcer is the one associated with the session with the highest rate of behavior.
  • Progressive schedule: response requirements for reinforcement are increased systematically independent of responding.  The practitioner gradually requires more responses to receive reinforcement until a breaking point is reached and responding declines. A progressive ratio increases the behavior that must be emitted before receiving reinforcement.  The reinforcers that produce the most behavior are considered the most reinforcing.

Behavior

Behavior is movement of an organism through space and time.  It must past the “dead man’s test” which is to say, “if a dead man can do it, it isn’t behavior.”  Behavior can only be done by a living organism and does not include movements due to physical pressures exerted upon that organism.  For example, someone getting blown over by the wind does not constitute behavior.  Someone rolling around and getting up because of the wind exerting force on their body does constitute behavior.

3 fundamental properties by which behavior can be measured (Johnston and Pennypacker 1993a)

  1. temporal locus – when in time that behavior occurs
  2. temporal extent – how long that behavior lasts
  3. repeatability – how many times that behavior happens over time (aka, rate of behavior)

 

Token Economy

Response Cost: Is a negative punishment that involves taking back tokens that have been earned.  It can be thought of as a fine for unwanted behavior which, hopefully, creates a decrease in the future possibility of this behavior occurring.

A token can serve as a Generalized Conditioned Reinforcer (GCR).  That is, it has been paired with many conditioned and unconditioned reinforcers.  GCRs are reinforcers that can be exchanged for other reinforcers and are less likely to be influenced by satiation and deprivation.  Money, for example, is a type of GCR and has little sensitivity to satiation.

Baseline Logic

Baseline logic looks at 3 dimensions of single-case design.  These are : prediction, verification and replication.

Prediction: What do you think will happen in the future
Verification: Showing that the dependent variables (DVs) would not change without intervention (independent variables: IVs).
Replication: Taking away the intervention, reintroducing it, and obtaining similar outcomes

Types of ABA Study Designs

Multiple baseline across participants is done by having multiple participants in the study and staggering when the intervention is applied.  At first baseline data is taken on all participants, and then participants are given treatment over time.

Alternating treatments design switches treatments back and forth, one at a time.  The treatments are “turned on / off” and compared on a graph over time.

Withdrawal / Reversal Design starts without the intervention for baseline, adds the intervention and then removes it.  The intervention should theoretically return to the baseline condition.

Changing Criterion design either gradually increases or decreases the behavior goals by continually over time changing the goal.

Behavior Intervention Plan (BIP)

Selecting interventions is based on:

  • what a behavioral assessment deems to be the problem behavior and its function as determined by data analysis
  • what the client needs in terms of a replacement behavior
  • the social significance of the change
  • evaluating the environment to see if it will produce reinforcement for the client after the intervention has been discontinued
  • keeping client’s preferences in mind

Good behavioral definitions are:

  • objective – referring only to observable characteristics
  • clear – no ambiguity
  • complete – giving examples of behavior and non-behavior