Multiple Regression

Question 1

purpose of correlations / simple linear regressions

Answer 1

examine relationship between a predictor and an outcome

Question 2

purpose of multiple regression

Answer 2

test unique association of multiple predictor variables, testing multiple associations at once

allows for categorical variables to be predictors

Question 3

unstandardized vs. standardized coefficients

Answer 3

unstandardized: amount that the dependent variable changes depends on the predictor variable

standardized: amount that DV changes is converted in terms of standard deviations

Question 4

statistical validity of regression coefficients

Answer 4

effect size: beta coefficients

confidence interval

statistical significance (p-value)

Question 5

assumptions of multiple regression

Answer 5

levels of measurement: outcome variable must be continuous

related pairs: no missing data

linearity

no multicollinearity: none of the predictor variables are correlated with each other

Question 6

how to report a regression coefficient

Answer 6

β = .xx, t(df) = x.xx, p = .xxx, 95% CI [.xx,.xx]

A greater number of friends significantly predicted more self-esteem when controlling for extraversion, β = .17, t(222) = 2.01, p = .045, 95% CI [.09, .25].