correlation

Question 1

what is correlation coefficient (r) and covariance?

Answer 1

single number that tells the strength and direction of correlation

sign tells us direction
- negative number = negative correlation
- positive number = positive correlation

value tells us strength
- range from -1 to +1
- closer to 0 = weaker
- closer to +/-1 = stronger

calculated from the covariance (how much two variables vary together)

Question 2

what is the effect size?

Answer 2

strength of correlation

can be interpreted by convention

small = +/-1
medium = +/-.3
large = +/- .5

Question 3

how do outliers effect correlations?

Answer 3

if an outlier has an extreme score but lies along the regression line, the correlation is stronger

if outlier is unusual in terms of combination, they will weaken the correlation

spearman’s row can be used instead of Pearson’s R to eliminate the influence of extreme values