Unit 531

library(tidyverse)
library(janitor)
library(haven)
library(broom)
library(foreign)
install.packages("haven")


healthdata <- read_spss("Health_LISS_Core_Study_Wave_12_2020_data_plus_background_small.sav")
healthdata %>% View()
healthdata %>%
  ggplot() +
  geom_histogram(aes(x = nettoink))

# filter income
healthdata <- healthdata %>%
  filter(nettoink <= 15000)

# scatterplot
healthdata %>%
  ggplot(aes(x = nettoink, y = BMI)) +
  geom_point() +
  geom_smooth(method = "lm", se = F)

# linear regression
model <- healthdata %>%
  lm(BMI ~ nettoink, data = .)

model %>% tidy()
summary(model)

# 95% confidence interval
confint(model, 'nettoink', level=0.95) %>%
  as.data.frame() %>%
  mutate_if(is.numeric, round, 6)