This preview shows page 1. Sign up to view the full content.
Unformatted text preview: you want to have a lower chance of a Type I error.
The ANOVA summary table
Extending ANOVA to Factorial designs
systematic variance broken up into subparts, for each of the main
effects and for the interaction terms
Don’t worry about calculation of these
You may see a more extensive ANOVA table, e.g.
How one would talk about the results?
Main effect for A, No main effect for B, Significant
Followup / Post hoc Tests
When there are more than two levels of an IV, you don’t know
which was the different one.
post hoc tests or multiple comparison tests
Used AFTER a significant F was found overall.
Between vs. Within-subjects ANOVAs
Just like before. Within-subject designs are more ‘powerful’
because they minimize error variance
Analysis of these requires special consideration (like the paired-t
test rather than the std t-test).
Multivariate ANOVA or MANOVA
Used when more than one DV.
Problem is inflation of Type I error when you have multiple DVs
you are analyzing. If 5 DVs you have a higher chance of one of
them showing up significant.
Nonexperimental use of ANOVAs
This is fine.
Be aware that just because ANOVA is used it doesn’t mean it’s a
equal variance between conditions
Single-subject research methods
Why? What’s the problem with group research? 5
Hides data - summarizing across subjects hides individual
The “Average individual” may not be similar to any subject (e.g.
Note, this does not preclude averaging within an individual.
Also, average group behavior can answer some questions about
how group on the average will perform - you just can’t address
individual behavior without further evidence
Simulation study of blocking - backward blocking and forward
Pigeon use of kinematic depth cues
Classic learning curve
Criticisms of group designs
The ‘averaging’ problem just mentioned
Error variance includes individual differences (intersubject
variance) as well as differences within an individual (intra
nomothetic (NA-MA-Thetic)- establish general principles and
idiographic (ID-E-A-Graphic) - seek to describe, analyze and
compare individual subject behavior.
Doesn’t imply only a single subject in the study
typically from 3 to 8, but analyzed and discussed as
Can use statistics (including inferential!!!) as well as graphical
Typically done under the rubric of Experimental Analysis of
Journal of the Experimental Analysis of Behavior
Journal of the Applied Analysis of Behavior
Analysis of single-subject data
The search for a STABLE baseline is...
View Full Document
This note was uploaded on 06/30/2012 for the course PSY 211 taught by Professor Chance during the Spring '11 term at University of Phoenix.
- Spring '11