Chapter 11: Quasi-Experimental Research Flashcards
Quasi-Experimental Designs
Experiments that study the effect of an IV in settings that cannot be controlled (but causal inference is hard to control, so internal validity is low)
- ie: One-group pretest-posttest design,
Quasi-Experimental Research: Key Points
- Used when true experiments are not possible
- Internal validity is lower (but can be built up in models like Control series design)
- Better for answering real world questions when done right
Types of Quasi-Experimental Designs
Not a True Experimental Design:
- One-Group Posttest-Only Design
- One-Group Pretest-Posttest Design
- Nonequivalent Control Group Design
- Nonequivalent Control Group Pretest-Posttest Design
- Interrupted Time Series Design & Control Series Design
One-Group Pretest-Posttest Design
Advantage: Help measure how much change or growth occurred between the pretest and posttest.
Disadvantage: Threats to Internal Validity
Threats to Internal Validity
History: outside event(s) that occur between the first and second measurement that are not part of the study, but may affect results
Maturation: The natural maturing of someone.
Testing effect: taking the pretest changes behavior without any effect on manipulation
Instrument decay: changes in measurement instruments (ie: humans) effects results
Regression toward the mean: extreme scores on a variable tend to be closer to the mean when a second measurement is made
Nonequivalent Control Group Design:
Two unequal groups participate in the different experimental groups, and there is no pretest.
Advantages: none…
Disadvantages: The differences between groups become a Confounding Variable, and thus a threat to Internal Validity
Nonequivalent Control Group Pretest-Posttest Design:
One of the most useful Quasi-experimental Designs!!
Advantages:
When a pretest is added, we determine whether or not the groups were equivalent.
Even if they were not, the prettiest-posttest allows us to look at changes in scored from the pretest to the posttest (not just at overall performance).
Example: if one group starts at 10 and ended at 20, and the other group started at 20 and ended at 30 – we can measure that there was a 10 pt increase for both groups
Propensity score matching
basically like a matched pairs design, after the fact.
Pairing similar individuals from a study the track results
Interrupted Time Series Design
Examine trends of DV over extended period of time (bother before and after introducing the IV)
Control Series Design
An extension of the interrupted time series quasi-experimental design in which there is a comparison or control group.
Developmental Research Designs
Cross-sectional method
Longitudinal Method
Sequential Method
Developmental Research Designs: Cross-Sectional Method
People of different ages are studied at one point in time.
Ie: examining average typing speed of American = you might look at people who are 15, 30, 45, 65, and 80
Disadvantage: Cohort effects
Developmental Research Designs: Longitudinal Method
Studying the same group of people over at different points in time, for an extended amount of time.
Disadvantage: expensive, long, low rate of follow through
Developmental Research Designs: Sequential Method
A combination of cross-sectional & longitudinal designs
Begins with cross-sectional (study each group in the moment)
After cross sectional study, then follow participants longitudinally
ie: using age: first study self-esteem in different ages,
* Then study the development of self-esteem over the next decade
* Results: understanding the development of self-esteem over a lifetime (done in only 10 years!)
