Quasi Experiments 7.3

Question 1

Drawing causal conclusions

Answer 1

• To demonstrate causality you need to be able to show that differences in a DV are driven only by the different levels of the IV
• A design free of confounds
• Comparing equivalent groups

Question 2

Quasi-experimental design types

Answer 2

• Pre-experiments
• Nonequivalent group designs
• Interrupted time series designs

Question 3

One-shot case study / one-shot posttest only

Answer 3

•A single group is measured after a treatment or intervention
•Behavior compared to hypothetical control group that didn’t receive treatment
ex:•Measure a person’s mobility after an accident

Question 4

One-group pre- and posttest design

Answer 4

•Measure the DV once before and once after
the treatment
•A within subjects design
•Many temporal confounds possible

Question 5

Nonequivalent groups designs

Answer 5

•Similar to true experimental designs

except the “IV” is a subject variable

Question 6

Nonequivalent groups without pretest

Answer 6

•Also called static group comparison
•Measured only after the event of 
interest (no pretest)
•Often used when researchers want to 
know how an event influenced people on a DV, but they did not know an event would 
occur or there was no way to obtain 
pretest data

Example:
Measure anxiety in individuals who experienced a
natural disaster

Question 7

Interrupted time series designs

Answer 7

Basic Interrupted time-series design: analysis of single interrupted time
series data that is hypothesized to be affected by an intervention

Question 8

(Basic) Interrupted time series design

Answer 8

Take multiple measurements before and after a treatment/intervention
- observing trends 
in a time series more 
informative than just 
pretest-posttest scores

Question 9

Interrupted time-series with non-equivalent

groups (multiple interrupted time series)

Answer 9

Like an interrupted time series, but it includes a non-equivalent control
group to provide more context for understanding the data.

Question 10

Interrupted time-series with switching

replications

Answer 10

•The same treatment or program is implemented in two different
locations but at two different points in time.
•No control group
•It has the benefit of a built-in replication

Question 11

Review of different types of quasi experiments

Answer 11

• Pre-experiments
• One-shot case study / one-shot posttest only
• One-group pretest-posttest
• Nonequivalent group designs
• With pretest
• Without pretest

•Time series
•Interrupted time series
•Interrupted time-series design with non-equivalent groups (multiple time-
series)
•Interrupted time series with switching replications