Chapter 8: Between subjects design

Question 1

What are the two sources of variance in a between subject design?

Answer 1

Systematic variance
Def→ differences among the means of different treatment groups
- difference in the DV between groups
- treatment effects + effects of chance factors (experimental error)
Non-Systematic Variance
Def→ Scores also vary within groups who are treated alike
- important source of errors
- individual differences occurring by chance
- experimental factors

Question 2

What is the F ratio formula for the between subject design?

Answer 2

F ratio= (treatement effects+experimental error)/experimental error

Question 3

How can we minimize within-group differences in between-subjects design?

Answer 3

Make sure all participants within a group are treated exactly the same
- Standardize the experimental procedures

Minimize individual differences:
- Hold extraneous variables constant or restrict the range

Use a large sample size
—>Random assignment has no effect on within-group variance

Question 4

What are the solutions to avoid confounding variables in between subjects design?

Answer 4

Randomization→ randomly assigned to groups to ensure groups are as equal as possible before treatment
- spread differences evenly
- for assignment of conditions
—>different from random sampling (random selection of participants from population)

Free Random Assignment→ Coin toss to ensure participants are assigned to groups solely on the basis of chance
- theoretically should lead to equality but not guarantee
- groups might not perfectly match but differ only randomly (pretty small)
- need big sample

Restricted random Assignment→ group assignment process is limited to ensure predetermined characteristics
- if want the group to be equal in size

Matching→ matched on critical variables
- ex: intelligence, gender…
- Match subjects on pre-existing differences
- can match across blocks→ ex: Age blocks (9-11/ 12-16…)
—>can be too costly to measure the variable

Holding Variables Constant
- ex: if think gender can be a confounding variable→ can use only women
- can also simply restrict the range of values
- ex: IQ range
—>but impact on external validity

Question 5

What are the threats to internal validity in a between subjects design?

Answer 5

Attrition
Problem if participants leave one group at higher rate
- groups not longer equivalent
—>Differential attrition

Communication between groups
Diffusion→ Treatment effects spread from one condition to another condition
- true effects might be masked
- have to be careful about group’s work

Resentful demoralization
Communication between participants can lead to perceived inequity
- ex: if one group receive course credit and another receive money
- observed differences between groups have alternate explanation

Question 6

What are the advantages of a between subjects design?

Answer 6

Simple design→ each score is independent
Clean and uncontaminated
- no carryover effects, no fatigue or boredom
Take less time for each participant
Causality can be established

Question 7

What are the disadvantages of a between subjects design?

Answer 7

Requires many participants
- can be difficult for special population
- ex: postpartum depression

Groups must be equivalent before the manipulation→ individual differences
- can become a confounding variable or produce high variability in the scores

Generalization can be hard
- holding subjects constant on extraneous variables reduces their representation in pop

Assignement bias and experimenter expectancy
- solutions is single (participants or experimenters) or double blindness
- Data analyst can also be blind to conditions of participants