7: Correlation Flashcards
What is bivariate correlation?
Examines relationship between two variables.
What is a strong correlation?
0.9 - 0.7 (+/-)
What is a moderate correlation?
0.6 - 0.4 (+/-)
What is a weak correlation?
0.3 - 0.1 (+/-)
Pearson’s r assumptions
Normal distribution, sensitive to outliers, and sensitive to range restrictions.
How is covariance calculated?
- For each datapoint, calculate the difference from the mean of x and from the mean of y.
- Multiply the differences.
- Sum them.
- Divide by N - 1.
What is a correlation coefficient?
Ratio of covariance to separate variances. Separate variances are obtained by multiplying the SD for x and y.
What is shared variance?
The amount of separate variances shared, expressed by r^2.
What is a partial correlation used for?
Examining the relationship between two variables whilst removing the influence of a third?
What are correlations useful for?
Predictions, validity (concurrent), reliability (test-retest, inter-rater), and theory verification (predictive validity).