bivariate correlational research

confidence interval

an index of how much we expect a sample to vary (uncertainty). a function of sample size. (bigger sample, smaller ci)

type 1 error

false positive. null hypothesis is incorrectly rejected leading to false conclusion of association. solution: replication

effect size

how practically significant the difference between 2 group averages is. (larger =more important, smaller = less important)

cohen’s d

measures effect size - difference between 2 group averages (means)

post hoc, ergo proper hoc

after that, therefore, because of that. fallacy that describes making a causal inference from correlational data

bivariate correlation

association that involves 2 variables

when to use a scatterplot (in bivariate correlation)

when both variables are quantitative

when to use a bar graph (in bivariate correlation)

when 1 variable is quantitative and one is categorical