fitting and interpreting models

how to write a linear regression model

linear_reg()

how to set model fitting engine to linear model

set_engine(“lm”)

formula syntax in R

• predates tidyverse

• provide formula then the data

std.error

uncertainty around estimates

slope interpretation structure

for each additional [explanatory variable + unit], the [response variable] is expected to be higher, on average, by [slope + unit]

least squares regression

• regression line minimizes the sum of squared residuals

• squared → get rid of neg so they don’t add up to 0 & more weight to residuals that are bigger

• goes thru center of mass point

center of mass point

• average of Xs

• average of Ys

correlation coefficient

measures the strength of the linear association two numerical variables

dummy variable

• encoded when categorical explanatory variable has many levels

• each coefficient describes the expected difference btwn heights in that particular level compared to baseline