PSYC300 Exam 3: Factorial Design

Factor

An explanatory variable (synonym for IV/PV)

Factorial design

Research design that includes 2 or more factors, e.g., 2×3×2—3 factors with 7 conditions

Factorial design disadvantages

-more complex
-harder to control all variables (lower internal validity)
-need more participants to have enough power

Factorial design advantages

-more efficient than doing separate experiments to test the effect of each IV
-more external validity
-allows for testing of main effects & interactions

Main effect

Mean differences among the levels/conditions of one factor

There is the same number of main effects as __

Factors

Interaction

One factor has a direct influence on the effect of another factor

T/F: Independent variables have interactions

False

One drug exaggerating the effect of another drug is an example of what?

Interaction

How can you tell if an interaction is present?

The lines denoting the difference between the main effects aren’t parallel

The two-factor study allows researchers to evaluate three separate sets of mean differences, which are…

Mean differences from the main effects of factor 1, 2, and interaction between factors

What are the different types of factorial designs?

-between
-within
-mixed
-combined

Between factorial design

All factors manipulated between participants

Within factorial design

All factors manipulated within participants

Mixed factorial design

At least one factor manipulated between & within participants

Combined factorial design

One factor is experimental and another isn’t; may or may not be mixed

__ depends on the level of measurement for the DV—continuous or discrete?

Statistical analysis

Statistical analysis also depends on what?

-level of measurement of factors (continuous/discrete)
-what the factors are (between, within, etc.)

What statistical analysis do we use for a between subjects design with a continuous DV & discrete factors?

N-way ANOVA

What statistical analysis do we use for a between subjects design with a continuous DV & continuous factors?

ANCOVA

What statistical analysis do we use for a within subjects design with a continuous DV & discrete factors?

N-way repeated measures ANOVA

What statistical analysis do we use for a within subjects design with a continuous DV & continuous factors?

Repeated measures ANCOVA

What statistical analysis do we use for a mixed design with a continuous DV?

Repeated measures N-way ANOVA

What statistical analysis do we use for a combined/non-manipulated design with a continuous DV?

-between: N-way ANOVA/ANCOVA
-within/matched: repeated measures N-way ANOVA/ANCOVA

When do we use ANOVA?

With discrete factors

When do we use ANCOVA?

With continuous factors

What do you do if the main effects are significant?

Post-hoc analyses

Post-hoc analyses

-Bonferroni corrections (conservative; reduced power; higher threshold for rejection—smaller alpha)
-Tukey test (honest significant test)—groups have equal variance