Chapter 9: Alternating Treatment Designs

Learning Objectives

• Describe procedures for and uses of alternating treatment design without baseline.
• Interpret data from alternating treatment design.
• Describe procedures for and uses of alternating treatment design with baseline.
• Understand alternating treatment design with baseline and final treatment phase.
• Discuss principles of prediction, verification, and replication in alternating treatment designs.
• Identify advantages and disadvantages of alternating treatment designs.

Alternating Treatment Designs Without Baseline

• Purpose: Compare the effects of two or more independent variables (treatments/interventions) on the same behavior.
• Usefulness: Effective for testing multiple intervention procedures to determine the most effective.
• Also Known As: Multi-element design.
• Characteristics:
  • Requires rapid alternation of treatments with measurement of their effects on a single target behavior.
  • Treatments can be alternated within the same session, across different times of the day, or across different days.
  • Helps avoid order effects.

Key Points in Alternating Treatment Design

1. Counterbalanced Presentation:
   • Treatments should be presented in a counterbalanced manner.
   • For three treatments (A, B, C), they can be presented randomly or in blocks ensuring each block is presented at least once.
2. Discrimination Among Treatments:
   • Subjects should distinguish between treatment conditions.
   • Treatments should be sufficiently different and may require cues such as verbal cues or signs.
3. Reversible Dependent Variable:
   • Behavior should be able to increase or decrease with different conditions.

Alternating Treatment Design Variants

• Three basic types:
  • Two include baseline data, and one does not.
• Baseline data may be omitted for ethical reasons (e.g., self-harm situations).
• No Treatment Phase: Can serve as a baseline condition in alternating treatments.
• Multiple Treatment Interference: Possible effect of one condition affecting others.

Case Study: Language Interventions in Children with Autism

• Purpose: Compare effectiveness of three natural language interventions.
• Design Used: Alternating treatment design without baseline.
• Dependent Variables:
  • Leon: Production of single words.
  • Griffin: Production of noun-verb and noun-adjective combinations.
• Independent Variables:
  • Responsive Interaction
  • Milieu Teaching
  • Combination of both
• Findings:
  • Milieu teaching and combination were more effective for Leon.
  • Responsive interaction improved over time for Griffin.
• Limitations:
  • Lack of baseline limits conclusions.
  • Multiple subjects and close proximity of conditions may cause interference.

Interpreting Data from Alternating Treatment Designs

• Effectiveness Determination:
  • Data paths of different treatments should be separate except at the beginning.
  • Calculate Percent of Non-overlapping Data (PND).
  • Compare data points from one condition to another sequentially.
  • Ideally, 100% PND indicates clear superiority.
• Example Analysis:
  • Condition C was superior 83% of the time compared to condition B.
  • Both conditions B and C were superior to no treatment (baseline) with 100% PND.