SPED 8013 | Chapter 19: Shaping

What is Shaping?

  • A process in which one
    • Systematically and differentially reinforces
    • Successive approximations toward a terminal behavior
  • Used to help learners acquire new behaviors

Differential Reinforcement

  • Reinforcement is provided for responses that share a predetermined dimension or quality, and reinforcement is withheld for responses that do not demonstrate that quality
    • i.e.,
      • Some members of a response class are reinforced (responses that are successively closer to the terminal behavior)
      • Other members of that response class (responses that are not closer to the terminal behavior)

Response Differentiation

  • Involves applying differential reinforcement consistently within a response class
  • Dual effects result in a new response class composed primarily of responses sharing the characteristics of the previously reinforced subclass

 Successive Approximations

  • Gradually changing the criterion for reinforcement during shaping that results in a succession of new response classes, each one closer in form to the terminal behavior than the response class it replaces

Shaping Different Dimensions of Performance

  • Topography
    • Form of the behavior
  • Frequency
    • Number of responses per unit
  • Latency
    • Time between onset of an antecedent stimulus and the occurrence of the behavior
  • Duration
    • Total elapsed time for the occurrence of the behavior
  • Amplitude
    • Magnitude

Shaping Diagrammed

  • Assume we want to teach a child to turn on the cold water tap  in order to get a drink of water
  • Assume the child already walks to the sink and looks at it when he/she is thirsty
  • Shaping might proceed like this:

  • Shaping Step 1

  • Shaping Step 2

  • Shaping Step 3

  • Shaping Step 4

Shaping Across and Within Response Topographies

  • Across response topographies
    • Topography of behavior changes during shaping
    • Behaviors are still members of the same response class
  • Within response topographies
    • Topography of behavior remains constant
    • Another measurable dimension of behavior is changed (e.g. duration of the behavior)

Positive Aspects of Shaping

  • Teaches new behaviors
  • Terminal behavior is always in sight
  • Frequent reinforcement
  • A positive approach to teaching
  • Can be combined with other procedures, such as chaining

Limitations of Shaping

  • Can be time consuming –many approximations may be necessary
  • Progress is not always linear and may be erratic
  • Requires a skillful and attentive trainer, who can recognize subtly closer approximations
  • Can be misapplied (problem or harmful behaviors can be accidentally shaped)

Shaping vs. Fading

  • Both change behavior gradually
    • Shaping – the antecedent stimulus stays the same, while the responses progressively become more differentiated
    • Stimulus fading – the antecedent stimulus changes gradually, while the response stays essentially the same

Increasing Efficiency of Shaping

  • Combine with a discriminative stimulus (e.g. a prompt)
    • Verbal cues
    • Physical guidance
    • Models
  • Any contrived prompt should later be faded

Guidelines for Implementing Shaping

  • Consider nature of behavior to be learned and resources available
    • How far away is current performance from terminal behavior?
      • This might influence how long shaping will take
    • What is the availability of staff and other resources?
      • Remember, this is a labor intensive procedure
  • Select the Terminal Behavior
    • The ultimate criterion for selecting a behavior for change:
      • How will the behavior change contribute to the learner’s independence in gaining reinforcement?
    • Define the terminal behavior precisely
      • Then you’ll know when the behavior has occurred (or when it hasn’t)
  • Determine Criterion for Success
    • How accurate, fast, intense, or durable must the behavior be? Under what conditions should it be performed?
    • Establish norms by:
      • Consulting literature
      • Observing similar peer groups
  • Analyze the Response Class
    • Identify the approximations that might be emitted during training
      • Trainer is in a better position to “stay ahead of his/her subject”
    • Can be done by:
      • Consulting experts in the field
      • Use published literature
      • Use videotape of peers to analyze components of a behavior
      • Perform the target behavior yourself
  • Identify the First Behavior to Reinforce
    • Behavior should already occur at some minimum frequency
    • Behavior should be a member of the targeted response class
  • Eliminate Interfering or Extraneous Stimuli
    • Eliminate distractions during training
  • Proceed in Gradual Stages
    • Be prepared for decrements in performance when you change criterion for reinforcement
  • Limit the Number of Approximations at Each Level
    • Lest the behavior become too firmly established and resistant to extinction
  • Continue Reinforcement When the Terminal Behavior is Achieved
    • The behavior will be lost if the terminal response is not reinforced