Lecture 9 Psyc 300A

1
Lecture 9Psyc 300A
2
Correlational Studies

• Why we do them
• Ethical limits on experiments and participant
  variables
• Often generalize well to other situations and
  people (external validity)
• Two major purposes
• Finding relationships
• Remember, cant get at causality. Why?
• Making predictions
• Criterion variable
• Predictor variable
• These studies require design considerations like
• Operational definitions
• Appropriate sampling

3
Correlation coefficients

• Comparing pairs of scores
• Scatterplots
• Correlation coefficent
• Ranges from -1.0 to 1.0
• Direction of relationship
• Strength of relationship
• Pearson product-moment correlation is for
  interval and ratio data
• Potential problems
• Nonlinearity
• Restriction of range

4
Scatterplot
5
Return to Experiments

• Logic of the experiment (IV and DV)
• Review of
• Extraneous variable
• Confounding variable
• Difference between the two

6
Properties of Studies

• Internal Validity
• Extent to which observed relationships in a study
  (scores) reflect relationships between
  hypothetical variables.
• External Validity
• Extent to which the results of a study can
  generalize to other people and settings outside
  the study

7
Designing Good IVs

• Strong vs weak manipulations
• Difference between levels of IVs must be big
  enough to see effects on DV
• Examples of strong manipulations
• Manipulation checks
• A measurement, separate from the DV, to see if IV
  had its intended effect

8
Designing Good DVs

• Sensitive measure
• A measure that is able to detect subtle
  differences in behavior
• Restriction of range
• Floor effects when task/test is too difficult,
  most scores will approximate the lowest possible
  score
• Ceiling effects when task/test is too easy, most
  scores will approximate the highest possible score