Lecture 7b: Quasi-Experimental Designs

Question 1

What are quasi-experimental designs?

Answer 1

subtype of non-experimental designs and resemble experimental designs but lack important features that allow for a causal inference
- lower internal validity than true experimental designs

Question 2

What are the two different types of quasi-experimental designs?

Answer 2

• one group designs
• non-equivalent groups designs

Question 3

What are one group designs?

Answer 3

no comparison group included in design, IV has only one condition/level
- equivalent to within-subjects design for a true experimental design

Question 4

What are non-equivalent groups designs?

Answer 4

includes comparison group(s) but participants are not randomly assigned to IV levels/conditions to create groups (using pre-existing groups)
- equivalent to between-subjects design for a true experimental design

Question 5

What are the two types of one-group designs?

Answer 5

• one-group posttest-only design
• one-group pretest-posttest design

Question 6

What is the one-group posttest-only design?

Answer 6

• weakest type of research design
• low internal validity
• descriptive statements only
• can’t make comparisons

Question 7

What is the one-group pretest-posttest design?

Answer 7

• can make comparisons about any changes in scores
• still don’t know whether the condition is more effective than others
• can’t rule out other explanations/confounding variables

Question 8

When are non-equivalent groups designs used?

Answer 8

when not possible for either practical or ethical reasons for the researcher to randomly assignment individual participants to different groups so we have to use pre-existing groups

Question 9

What are the two types of non-equivalent groups designs?

Answer 9

• non-equivalent control group design
• non-equivalent control group pretest-posttest design

Question 10

What is non-equivalent control group design?

Answer 10

• increase in internal validity because can make comparisons between conditions
• more confidence in making conclusion that one condition is better than the otehr
• can’t rule out confounding variables
• higher risk of selection differences because groups are pre-existing and not created through random selection - limits causal inference

Question 11

What is non-equivalent control group pretest-posttest design?

Answer 11

• establish pre-existing differences through pretest
• can compare greater increase in pre- and post- test by comparing differences
• remove some possibility of an alternative explanation
• now have more information about how the groups compare on RM knowledge but still some risk of alternative explanation for results – limits internal validity

Question 12

What are the advantages of using quasi-experimental designs?

Answer 12

• better internal validity than other non-experimental designs (especially non-equivalent groups designs)
• sometimes the only way to study a certain topic

Question 13

What are disadvantages of using quasi-experimental designs?

Answer 13

• poorer internal validity than true experimental designs

Question 14

Why would you use a quasi-experimental design?

Answer 14

• control/comparison group or random assignment groups is not possible
• control/comparison group or random assignment to groups would be unethical
• can have higher external validity than experimental designs
• researcher’s goal is to describe or predict behaviour