Chapter 10: Complex Experimental Designs Flashcards
Better Than Average Effect (BTA or BEA)
Definition: The difference between
People tend to rate themselves more favorably than they rate other people.
When should you increase levels?
one IV, with 3+ levels
- If there is reason to predict a non-linear relationship
- In order to test a specific hypothesis
- The study is exploratory
Factorial Design: Definition
An experiment with more than one IV where all levels of each IV are combined with all levels of the other IV.
ex: 2x2 factorial design has 2 IVs, each with 2 levels. So, there would be 4 total conditions.
Why use a Factorial Design?
Allows investigation of the separate main effects and interaction of two or more independent variables
Types of Factorial Designs
- Between-Subjects Factorial Design
- Within-Subjects Factorial Design
- Mixed Factorial Design
Between-Subjects Factorial Design
Each participant goes through only one combination of levels of variables
- All variables are manipulated between-subjects.
Within-Subjects Factorial Design
All participants go through all levels of every variable.
*All variables are manipulated within-subjects.
Mixed Factorial Design
At least one variable is manipulated via within-subjects and at least one variable is manipulated via between-subjects.
Factorial Design - Effects
- Simple (main) Effect
- Main Effect
- Interaction
Simple (main) Effect
The effect of one IV at a specific level of another IV
If there is NOT interaction, simple effects will not be different from each other (usually)
Main Effect
The direct effect of an IV on a DV
Main effects may not be to interpret when an interaction is present.
Interaction
When the effect of one IV changes based on the level of another IV
Interaction = a difference between differences
To Understand Interactions first…
- first looking at simple effects.
- Then compare the simple effects to each other.
- If they are not the same, there may be an interaction.
Describing an Interaction
a) State the effect at one level of an IV (one simple effect)
b) State the effect at the other level of an IV (other simple effect)
c) COMPARE the two statements (the two simple effects)
How many Effects In a 2x2 Factorial Design
- 4 simple effects – but typically only report 2 to illustrate a significant interaction
- 2 main effects – important to report both.
- 1 interaction – one interaction with two different “angles”
Assigning Participants to Conditions: Within-subjects design
The same person participates in every condition
Within-Subjects 2x2 Factorial Design:
- Four total conditions, but only one group of participants needed.
- If each condition needs 10 people, you would only need 10 different participants total. n=10
Assigning Participants to Conditions: Between-subjects design
Different groups of participants are assigned to each of the conditions.
Ie: Between-Subjects 2x2 Factorial Design:
- Four total conditions. Thus, four different participant groups.
- If each condition needs 10 people, you would need 40 different participants total. n=40
Assigning Participants to Conditions: Mixed Factorial Design
Experiment includes both Between and Within groups designs, so one IV would need different participants for each condition, and the other would only need one group of participants
Mixed 2x2 Factorial Design: (IV-a = between, IV-b = within)
- Four conditions total. 2 conditions will require separate participant groups, while 2 conditions will only need one.
- If each group needs 10 participants, IV-a would need a total of 20 participants & IV-b would need a total of 10
- Grand total of n=30
Interaction “angles”
2 ways to look at one Interaction:
- Does the effect of IV-a depend on the level of IV-b?
- Does the effect of IV-b depend on the level of IV-a?
Which “angle” is best?
Use the angle that directly address’ the research question
But, both help understand the interaction.
