Chp 3 Flashcards
Group experimental designs
Evaluate if the behavior of a treatment group (independent variable ON) is statistically significantly different from that of a control group (independent variable OFF).
Used in clinical behavior analysis
Weaknesses of group experimental designs
- When the independent variable is therapy, no one wants to be assigned to control group
- Focusing on group behavior means we aren’t studying individual behavior
- The behavior of the treatment and control group will differ simply bc the people assigned to the 2 groups are different
- Reliance on inferential stats to evaluate if independent variable changed behavior
Single subject experimental design
Exposes individuals to baseline (independent variable OFF) and experimental (independent variable ON) phases to determine if the independent variable systematically and reliable changes behavior.
Internal validity
When an experiment provides clear evidence that a functional relation exists between the independent variable and behavior change
Confounds
Variables that influence behavior within an experiment but aren’t controlled by the researcher
Types of single subject experiments
- Comparison design (A-B)
- Reversal design (A-B-A)
- Alternating treatments design
- Multiple baseline designs
Comparison design (A-B)
Simplest single subject design that arranges a baseline (A) phase (independent variable OFF), and experimental (B) phase (independent variable ON). aka quasi-experiment
Baseline phase
Phase allows for accurate predictions of future behaviors. During Baseline the independent variable is OFF so researchers can base research off the baseline.
Behavior is stable when
Over repeated observations, there is little bounce (variability) and no systemic trend
Reversal design (A-B-A)
The individual’s behavior is evaluated in repeatedly alternating baseline (A) and experimental (B) phases. Better equipped to rule out confounds
Alternating treatments design
The independent variable is turned ON and OFF rapidly to evaluate if this systematically and repeatedly changes behavior. Used to evaluate relations between hs behavior and 1+ independent variables.
Multiple baseline design
Evaluates the functional relation between an independent variable and behavior by conducting a series of time-staggered A-B comparisons either across behaviors, across situations, or across individuals.
4 defining features of single subject experimental designs
- The focus is on the behavior of individuals not groups
- Each subject experiences the baseline and experimental (intervention) phases.
- Behavior is measured repeatedly in each phase until confident predictions about behavior may be made
- Internal validity is assessed through replication and evaluating the functional role of confounded variables
3 kinds of replication in single subject experiments
- Within individual replication
- Across-individual rep
- Rep across labs or clinics
Objective decision-rules in answering ‘did behavior change?’
- Inferential statistics
- Visual analysis of behavior
Inferential statistics
Scientists conduct experiment and enter data into a computer and computer uses stats to report the results.
problems: the need for greater transparency (asking scientists exactly how the experiment was conducted), it requires a lot of participants, and it isn’t used w time-series data
Visual analysis
Involves looking at a graph of time-series single subject behavior to evaluate if a convincing change occurred when the independent variable was introduced/removed
Advantages to visual analysis
- It requires methods for producing large changes in behavior, obvious changes
- It recognizes the inherent subjectivity of the science enterprise
2 patterns of change
- Trend (a systematic change in behavior over time)
- Level (the prevalence of the behavior during the stable portion of a phase/condition)
Relation between bounce & trend/level
Minimal bounce in the graph makes the trend and level obvious.
The trend and level change must be large relative to the bounce)
Guidelines for visual analysis
- Draw a trend arrow through the baseline data to predict what will happen if the independent variable is never turned ON
- Evaluate if behavior in baseline is too variable (bouncy) to have confidence in the prediction of the trend arrow
- Draw trend or level lines through the intervention data. Evaluate if there is a convincing change in trend or level (whichever change is of interest)
