Interventions
B Design
The B design is a single phase (intervention only) design.
0The most frequently used single subject design.
0Only B design does not allow for evaluation of pre-intervention status, we do not know if there is a relationship between intervention and outcome.
Conceptual Definition
Definition from dictionary
Levels of measurement
Nominal: categories
0Ordinal: categories in a certain order
0Interval: equal intervals
0Scale: equal intervals and 0 point
AB Design
Two phase design consisting of a no-intervention baseline phase (A) and an intervention phase (B).
0The existence of a baseline allows for the establishment of a relationship between intervention and outcome.
0Susceptible to uncontrolled influences of extraneous variables (the non-spuriousness issue in causal inference), especially the history threat to internal validity.
Operational Definition
something that you can measure
Characteristics of measurements
Validity
•Reliability
•Measurement error
•Utility
•Directness
ABA Design
A three phase design: 1) No-intervention baseline phase (A), 2) Intervention phase (B), and 3) No-intervention withdrawal phase (A).
0Allows for evaluation of pre-intervention and intervention problem status.
0More reliable establishment of a relationship between intervention and outcome than in the AB design.
Ultimate goal
what client would like to do, be or prefer to happen after the intervention is complete.
Accuracy is...
Validity
BAB Design
Sometimes an individual’s behavior is so severe that the researcher cannot wait to establish a baseline and must begin with an intervention.
0In this case, a B-A-B design is used. The intervention is followed by a baseline followed by the intervention.
Intermediate goals
very specific versions of the ultimate goal
Consistency is...
Reliability
Multiple Baseline Design
Allows for evaluation across clients, situations, or problems.
0True experimental design in that it allows for causal inference.
0Useful for evaluating situations where an intervention would be likely to bring about enduring changes in the dependent variable.
Types of measurement error
Two types
•Random: by chance, normal fluctuations, cancel each other out
•Systematic: bias, average score is influenced into a particular direction