PSY 370 Lecture Notes - Lecture 16: Statistical Significance, Aspirin, Psy
Document Summary
Size of the correlation indicates the strength of the relationship. A correlation of -0. 75 is bigger than a correlation of 0. 50. A very small correlation means the variables are unrelated. We should expect a smaller number than the correlation (r) itself. In testing and measurement, how big should a correlation be in order to be considered meaningful. "statistical significance" of correlations depends on sample size. Lots of debate about significance testing in general. Even a very small correlation can have a big practical impact. Correlation between treatment and survival: r = . 20; r2 = . 04. This means 4% of the variance is explained. Correlation between daily aspirin consumption and reduced risk of death from heart attack: r = . 02. Correlation between alcohol consumption and aggressive behavior: r = . 23. Correlation between nearness to the equator and average daily temperature: r = . 60. Two different things should not be correlated at . 9! Funder & ozer (2019) suggest a rough guideline: