Factorial design

Question 1

What is a factorial design

Answer 1

experimental design used to study the effects of two or more independent variables (factors) on a dependant variable at the same time

Question 2

Who is Ronald Fisher

Answer 2

• Followed idea that When you’re experimenting with multiple factors (like temperature, time, noise, etc.), you can’t assume that one variable acts alone.
• The effect of one factor might depend on the level of another - this is called an interaction.
• So rather than testing one factor at a time (which ignores interactions), you should test combinations of factors — which is exactly what factorial designs do.

Question 3

What are the factors

Answer 3

independent variables in your experiment.
Example: Type of music, Time of day, Type of task

Question 4

What are the levels

Answer 4

different conditions within each factor.

Example:
For Music → Classical, Pop, None (3 levels)
For Time of day → Morning, Evening (2 levels)

Question 5

What’s a main effect

Answer 5

A main effect is the effect of one independent variable (factor) on the dependent variable, ignoring all other factors.
(averaged by the over the levels of the factors)
then the effect is the difference between the means

example:
Main effect of drink = Do people perform differently depending on the drink, on average, no matter what time of day?

Main effect of time = Do people perform differently depending on the time, on average, no matter what drink?

Question 6

What is an interaction

Answer 6

The difference between the levels
An interaction happens when the effect of one factor depends on the level of the other

Question 7

Plotting Main effects

Answer 7

• When slopes are parallel, the two main effects do not interact (or depend on each other for their values). Therefore, we would say the effects of coffee and stress on test scores are independent of each other.

Question 8

Plotting interactions

Answer 8

• No interaction = the two lines on a graph are roughly parallel
• Interaction = the lines cross or diverge = effect of one factor is not consistent across levels of the other

Question 8

What is a triple interaction

Answer 8

the interaction between 2 factors differs across the levels of the other factor

Question 9

Independent (Between-Subjects) Factorial Design

Answer 9

Each participant is in only one condition.

You need a different group for every combo of your factors.

Question 10

Within-Subjects Factorial Design

Answer 10

Every participant experiences every combination of conditions.

Fewer participants needed, but order effects can be an issue.

Question 11

Mixed Factorial Design

Answer 11

At least one factor is between-subjects, and at least one is within-subjects.

A mix of both: some things change between groups, others change within the same people.