6.1 Data & Packages & functions

• Objective: Summarize results in graphs instead of unreadable tables
• Statistical results: Stored in model objects
  • Inspect the model object with modelname$...
• Two approaches
  • Estimates models and extract estimates from model objects (and other characteristics) to visualize them
    • broom::tidy(): Use the tidy function to extract estimates etc.
    • broom::confint_tidy: Add confidence intervals
  • Use plotting functions that directly plot estimates from model objects
    • e.g., sjPlot package package

fit <- lm(Fertility ~ Catholic, data = swiss) # see ?swiss
# fit$... 
coef(fit)

## (Intercept)    Catholic 
##  64.4282621   0.1388857

tidy(fit)

term	estimate	std.error	statistic	p.value
(Intercept)	64.4282621	2.3050957	27.950363	0.0000000
Catholic	0.1388857	0.0395606	3.510712	0.0010285