9: Small N Designs Flashcards
Describe narrative case studies and what they are used for.
Intensive study of single person.
Provides rich information, can be initial basis for new theory/interventions, useful for discussing exceptions.
List four advantages to narrative case studies.
Source of ideas and hypotheses.
Source for developing new therapy techniques.
Can study rare phenomena.
Inform clinicians about side effects of treatment.
List five disadvantages to narrative case studies.
Limited generalizability.
Alternative explanations often available.
Many threats to validity.
Relies heavily on anecdotal information.
Non-representative cases often presented.
What are single-case experimental designs? What are they similar in general philosophy to?
Inferences made about effect of intervention by comparing different conditions presented to same subject overtime.
Repeated measures group designs.
Single-case experimental designs include what four essential components?
Continuous assessment: enables meaningful comparisons to be made.
Baseline assessment: multiple observations; provide info about behaviour of interest before intervention.
Stability of performance: relatively stable at baseline; high variability or trends in slope interfere with evaluation.
Multiple phases: inferences drawn based on patterns across phases.
List three limitations to single-case experimental designs.
Still just quasi-experimental.
Low internal validity; cannot exclude many competing explanations.
Low external validity.
Silverman, Wong, et al. (1998) conducted what study? Describe the design.
Cocaine abstinence - escalating reinforcement via monetary compensation.
Baseline, reinforcement, then withdrew reinforcement.
List three problems with A-B-A-(B) designs.
Not all treatment-related behaviours reversible.
May be ethical problems regarding withdrawing treatment.
Switching treatment “on and off” may have undesirable consequences (e.g., lack of trust).
Multiple baseline designs avoid what problem? What is a con?
Avoids ethical problem of having to reverse treatment effect.
Behaviours cannot be highly interrelated, otherwise treating one behaviour would be like treating the others.
What are the four steps to changing-criterion design?
- Take initial baseline measure.
- Apply treatment over series of trials.
- Each successive trial, set criterion threshold for defining a treatment response a little higher.
- Experimental control demonstrated through successive replications of change in the target behaviour.
In changing-criterion design, treatment trials must be long enough to allow for what two things?
New behaviour to emerge.
Demonstrate stability in new behaviour.
In changing-criterion design, magnitude of behaviour must be what?
Large enough to be observable, but small enough to be considered incremental.
Identify four challenges in single-case design.
Difficulty establishing stable baseline.
Variability in baseline data makes it hard to know when baseline trend has stabilized.
Interpretation of treatment efficacy can be ambiguous without statistics.
Behaviour may be reactive to measurement, thus impossible to obtain a valid measurement.
How is data analysis usually conducted in single-case design? What is the goal?
Generally without statistical tests, use visual inspection.
Identification of effects that are reliable and unlikely due to chance.
List the four criteria for visual inspection.
Mean: reflects shift in average performance from one phase to next in expected direction.
Level: difference between last day of previous condition and first day of new condition.
Slope (trend): direction or magnitude of behavior change during different phases.
Latency of change: how quickly the difference between two conditions becomes apparent; shorter is better.
