Week 7

Question 1

What is a quasi-experimental design?

Answer 1

A design in which an experimental procedure is applied but all extraneous variables are not controlled.
Fall between strong and weak experimental designs.
Does not include random assignment.

Question 2

What are the different types of quasi-experimental design?

Answer 2

1. Non-equivalent comparison group design.
2. Time-series design.
3. Regression discontinuity design.

Question 3

How do you rule out rival explanations in quasi-experiments?

Answer 3

1. Identification and study of plausible threats to internal validity (you are fucked if you don’t pick up on them).
2. Control by design - Pretests and control groups.
3. Coherent pattern matching - Being able to predict complex patterns that cannot be explained by other hypothesises.

Question 4

What is a non-equivalent comparison group design.

Answer 4

A design that includes pretesting of non-randomly sampled groups, assigning control groups and experimental groups, applying treatment and then posttest groups.

Question 5

What effects could be present if control and experimental groups increase in scores during a time series study?

Answer 5

Selection-maturation or selection-history.

Question 6

What effects could be present if the experimental group starts with higher scores than the control group at pretest with an increase in scores only in the experimental group?

Answer 6

Selection-history.

Question 7

What effects could be present if the experimental group starts with lower scores than the control group at pretest with only the experimental group scores increasing closer to the control group?

Answer 7

Selection-regression.

Question 8

What effects could be present if the experimental group pretests with lower scores than the control group but scores increase higher than the control (if the control group does not change)?

Answer 8

Likely none.

Question 9

What is equality and equity?

Answer 9

Equality - Same treatment regardless of recipient.

Equity - Treatment varies based on needs.

Question 10

What is ANCOVA?

Answer 10

Analysis of covariance.

Question 11

What is interrupted time series design?

Answer 11

A quasi-experimental design in which treatment effect is analysed by comparing patterns of pretest and posttest patterns (changes in level and slope) .

Question 12

What should you keep in mind when using interrupted time series design?

Answer 12

1. Delayed effects.
2. Temporary effects.
3. Gradual effects.

Question 13

What is regression discontinuity design?

Answer 13

Quasi-experimental design that assigns participants to groups based on scores for a variable.
Administers an experimental effect to a group, then looks for discontinuity in regression lines.

Question 14

What are single case designs?

Answer 14

A research design in which a single participant or group is used to investigate a treatment.

Question 15

What are the types of single case designs?

Answer 15

1. Time-series/ABA (Baseline, treatment, baseline).
2. Time-series/ABAB (Baseline, treatment, baseline, treatment).
3. Interaction design.
4. Multiple-baseline design.
5. Changing-criterion design.

Question 16

What is baseline?

Answer 16

Behaviour of participant in natural state/prior treatment.

Question 17

What is reversal?

Answer 17

Return to baseline after treatment.

Question 18

What are problems with ABA/ABAB designs and how do you fix them?

Answer 18

1. Ending on baseline is unacceptable for patient participants - fixed with ABAB.
2. Reversal may not occur - Fixed with multiple baselines.

Question 19

What are withdrawal and reversal designs?

Answer 19

Withdrawal design: Removal of treatment condition.

Reversal design: Apply treatment which produces reversal.

Question 20

What are interactive designs?

Answer 20

Time-series designs (ABA) with multiple treatments, can be used with multiple participants or a single participant.

Question 21

What is a multiple-baseline design?

Answer 21

A single-case design in which a treatment condition is applied to multiple individual participants in a staggered time interval.

Question 22

What is interdependence?

Answer 22

An issue that arises in multiple-baseline studies in which changes in one participant produce changes in remaining participants.

Question 23

What is changing-criterion design?

Answer 23

Starting with a baseline, applying treatment until criteria is met, change criteria, repeat.

Question 24

Why is a stable baseline important and what constitutes a stable baseline?

Answer 24

It allows for a stable reference point.

It must be: Little variability, stable, not influenced by participant reactivity.

Question 25

What should you take into consideration when doing a single-case designs?

Answer 25

1. Length of phases - Long enough for behaviour to stabilise.
2. Criteria change size - Large enough to notice change.
3. Number of phases - At least two, need enough to demonstrate effect.

Question 26

What is the criteria for evaluating change in single-case designs?

Answer 26

1. Experimental criterion - Repeatedly demonstrates behavioural change when treatment is applied (need lots of examples).
2. Therapeutic criterion - Demonstration that disorder has improved after treatment (can be placebo).
3. Social validation - Professional report that behaviour has significantly changed (can be biased).
4. Social comparison - Participant is compared to nondeviant peers (do them seem normal).
5. Subjective evaluation method - Others perceive change in participants behaviour.