Correlations Pt. 1

when to use the Pearson correlation statistic

• correlation statistic: the strength and direction of the relationship between two quantitative variables

  • example of measured quantitative variables:

    • age

    • time to finish a marathon in minutes

    • number of details you can remember about a story you just heard

    • number of games won in a season

    • number of alcoholic drinks per week

    • number of steps per day

• correlation statistics are only for linear relationships that can be modeled or “fit” with a straight line

  • the strength of the correlation can be visually estimated by how close the points are to a trend line

    • range -1 to 1

  • the direction is shown by the slope

    • positive/negative

• r is a “unitless” descriptive statistic

  • r does not reflect the original units of the variables

  • r is an effect size for the strength of the relationship

  • effect sizes are comparable across different variables and studies

steps to compute the Pearson correlation statistic

computing r

how to describe your correlation result in words

• association claims: assert that two variables are related to each other

  • the frequency of one variable is tied to or linked with the frequency of another

  • pearson correlation statistics support association claims