Correlation (Descriptive)

Correlation (Descriptive)

1. Correlation coefficient (e.g., r)

   • r quantifies the size and direction of the relationship between 2 variables
   • r is a descriptive line statistic, and can be affected by a number of factors

2. When there is a significant correlation between X and Y, all we can conclude is that X is "related" to Y

   • Correlations have got to make sense; that is, interpretable in theory

3. Spurious Correlations: Chance relationships between 2 variables,

4. Effect Size (r^2)

5. Assumptions

   • Independence of Observations: 1 person can only contribute 1 score
   • Bivariate Normality: The X and Y variables are normally distributed
   • A statistical test is valid if its assumptions about the data are met. The degree to which these assumptions are met will determine the degree to which the test is valid

Images and files in the map