Complex Designs

1
Complex Designs

• Psych 7
• February 27th

2
Outline

• Factorial Designs
• Main Effects
• Interactions

3
Factorial Designs

• Designs that contain 2 variables that are
  completely crossed
• The label a researcher gives a factorial design
  specifies both
• of (IVs)
• of of each IV
• 2 X 2
• 2 X 3
• 2 x 2 x 2

4
Experimental Designs

• Previously
• Single-factor designs
• One independent variable
• Now
• Two (or more) factor designs
• E.g., What is the relation between caffeine
  consumption, alcohol consumption, and reaction
  time

5
A 2x3 Design
No Caffeine 200 Mg 400 Mg
Alcohol
No Alcohol
DV How quickly Ps apply the brakes in a driving
simulation task
6
Benefits of Factorial Designs

• What can we learn from factorial designs that we
  cant learn from single-factor designs?
• Examine more complex, real-life situations
• Examine how each individual factor influences
  behavior
• Examine how factors influence each other

7
Main Effects

• Factorial designs allow researchers to answer
  questions about more than one IV at the same time
• Main Effects
• e.g., the effect of gender and frustration on
  aggression

8
Main Effects in a 2x2 Factorial Study

• Greenwald, et al. (1991)
• Examined the effects of expectations on memory
• Ps listened to one of two subliminal self-help
  tapes
• IV1 Message content (self-esteem, memory)
• IV2 Tape label (self-esteem, memory
• DV Perceived improvement in memory

9
Greenwald, et al. (1991) Results
2(label on tape self-esteem, memory) x 2(tape
content self-esteem, memory)
10
Greenwald, et al. (1991) Results Label Main
Effect
2(label on tape self-esteem, memory)
11
Greenwald, et al. (1991) Results
2(label on tape self-esteem, memory) x 2(tape
content self-esteem, memory)
12
Greenwald, et al. (1991) Results Tape Content
Main Effect
2(tape content self-esteem, memory)
13
Interactions

• It depends!
• The effect of an IV on a DV depends on the level
  of a second IV
• e.g., When forming impressions, when will we take
  the situation into account?

14
Interactions in a 2x2 Factorial Study

• Gilbert, Pelham Krull (1988)
• Ps watched a video and saw a woman behaving
  anxiously
• IV1 anxiety-provoking topic or relaxing topic
• IV2 cognitive load or not
• DV Judged level of anxiety

15
Gilbert, Pelham Krull (1988) Design
Level of Cognitive Load
Discussion Topics
Low (one task) High (two
tasks)
Anxiety Provoking
Relaxing
16
Gilbert, Pelham Krull (1988) Results
2(cognitive load low, high) x 2(topic anxious,
relaxing)
17
How to Recognize Main Effects and Interactions

• Main Effects
• Look at the table of means
• Look at the graph
• Interactions
• Look at the graph

18
Effect of Body Size and Gender on Yearly Income

• Quasi-experimental factorial design (more on
  these to come)
• IV1 Height
• Tall vs. Short
• IV2 Gender
• Male vs. Female
• Fully crossed design
• Tall Males, Tall Females, Short Males, Short
  Females
• DV Yearly Income

19

20
(No Transcript)
21
(No Transcript)
22
Tall
Short
23
Tall
Short
24
Tall
Short
25
Tall
Short
26
Tall
Short
27
How to Recognize Main Effects and Interactions,
cont.

• x if points are different, main effect for
• if points are different, main effect for
• If lines are parallel

• x no main effect for
• no main effect for
• Lines are not parallel

28
Interactions

• In a 2 X 2 design, can you interpret the main
  effects without interpreting the interactions?

29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
Interactions

• In a 2 X 2 design, can you interpret the main
  effects without interpreting the interactions?

33
Thursday

• Quasi-Experimental Designs
• Homework 2 due on March 6th
• You should now have all the info you need to
  complete this assignment