PSYC121_2021-22/Week_7_script_answers.R at ff15dfed06b8433583dee0cf6a6d6d4dc47cdad4

Cannot retrieve contributors at this time

56 lines (42 sloc) 1.74 KB

	library(tidyverse)
	options(scipen = 999) # this stops numbers appearing as scientific notation

	# TASK 1

	# read in the data
	data_w7 <- read_csv("uk_salary.csv")

	# quick look at the salary estimates
	data_w7 %>%
	ggplot() +
	geom_histogram(aes(x = uk_salary)) # EDIT THIS CODE FOR Q2

	# let's look at those unusual values
	data_w7 %>%
	arrange(desc(uk_salary)) # arrange the data according to descending values of uk_salary estimates

	# let's remove those really high values
	data_w7_f <- # note the assignment (<-): the result of this calculation will make a new object
	data_w7 %>% # with the data...
	filter(uk_salary < 200000) # EDIT THIS CODE FOR Q4

	# now let's plot the salary estimates again
	data_w7_f %>%
	ggplot()+
	geom_histogram(aes(x = uk_salary)) # EDIT THIS CODE FOR Q5

	# display the data as a boxplot
	data_w7_f%>%
	ggplot() +
	geom_boxplot(aes(y = uk_salary)) # EDIT THIS CODE FOR Q6

	# TASK 2

	# What is the mean estimate?
	mean(data_w7_f$uk_salary) # EDIT THIS CODE FOR Q7
	sd(data_w7_f$uk_salary) # you might also calculate the standard deviation

	# Let's check if the mean estimate is significantly different from the true salary
	t.test(x = data_w7_f$uk_salary, mu = 30000) # You will need to edit this for Q8

	# EXTRA
	# this is a "by hand" t statistic - you can add the column of salary estimates here and check with the result of t.test
	# check with staff if you're not sure
	t_val <- (mean(data_w7_f$uk_salary)-30000)/(sd(data_w7_f$uk_salary)/sqrt(length(data_w7_f$uk_salary)))

	# TASK 3

	# plot the uk_salary data using geom_density
	# EDIT THIS CODE FOR Q11-14
	data_w7_f %>%
	ggplot() +
	geom_density(aes(x = uk_salary, fill = home_location),
	alpha = .5) +
	scale_fill_manual(values = c("darkgreen", "darkblue", "darkred"))

Unstar this repository?

PSYC121_2021-22/docs/files/Week_7/Week_7_script_answers.R