Flashcards in KL Chapter 22 Quasi-experimental and N=1 Designs Deck (15):
When does one use quasi-experimental design?
When one or more prerequisites to a true experiment is missing
Nonequivalent control group design
no clear assurance that the control and experimental groups are equivalent
8 ‘interpretations’ of Nonequivalent control group design
1. no-treatment control group
2. nonequivalent dependent variables design
3. removed treatment group design
4. repeated treatment design
5. reversed treatment nonequivalent control group design
6. cohort design
7. posttest only design
8. regression continuity design
threat to internal validity. occurs when the two groups are different at pretest and one of the groups experiences greater change. The posttest difference cannot be attributed to the treatment.
threat of instrumentation
indicates that the measure of the dependent variable may be more sensitive at certain levels than others
threat of statistical regression
indicates participants were chosen based on their extreme scores.
indicates that something other than the independent variable affects one group, but not another.
Longitudinal time design
interrupted time series design
Multiple Time series design has what advantage?
eliminates the history threat to internal validity
What are Single subject experimental designs (N=1) ?
are an extension of the interrupted time series design, but with 1 unit (participant).
Advantages of single subject experimental designs?
actual performance of participant shown, as opposed to misleading group graphs. Large samples can be stat. sig. with very small actual differences
Disadvantages of single subject experimental designs?
problems with external validity, these studies are sensitive to researcher bias
Purpose of baseline in single subject design?
necessary to determine if the intervention had an effect
o Multiple baseline studies include designs...
across behaviors, across participants, and across environments