College Rankings and State Characteristics
We will work with the tidyverse package as usual. You may also want to use viridis.
library(tidyverse)
library(viridis)The U.S. News rankings are an influential, but controversial metric that influences the college application process.
A brief description of the data sets for this lab and how they are related to each other is provided below.
The natunivs data set contains all schools in the National Universities category ranked 50 or above in the current (2022) rankings. Data on school rankings comes from Andy Reiter, with several blank values filled in by the professor. Observations are uniquely identified by school.
The variables in this data set are:
school: The name of the college or university.state: The state in which the college or university is located.rank_2022: The school’s rank in the 2022 issue.rank_2021: The school’s rank in the 2021 issue.natuniv_slac: A variable identifying the type of school.
The slacs data set contains contains all schools in the National Liberal Arts Colleges category ranked 50 or above in the current (2022) rankings. Data on school rankings comes from Andy Reiter. Observations are uniquely identified by school.
The variables in this data set are the same as above:
school: The name of the college or university.state: The state in which the college or university is located.rank_2022: The school’s rank in the 2022 issue.rank_2021: The school’s rank in the 2021 issue.natuniv_slac: A variable identifying the type of school.
The state_data data set contains three variables related to the characteristics of a state:
abbrev: The state’s abbreviation.2020_pop: The state’s population in the 2020 census.2020economicoutlook: A state’s economic outlook score (1-50)
State economic outlook scores come from richstatespoorstates and are compiled from 15 markers of economic stability, including minimum wage, state tax burden and employment levels. The 2020 population data comes from the US Census.
Looking at this data
- Let’s start by creating a data set that includes information from all three data sets.
First, join the
slacsdata set to thenatunivsdata set. The goal is to combine these data sets in such as a way so that the new data set has the same number of rows and unique columns as each of the individual data sets. Call this new data setfull_data.Next, use a join to add the columns from the
state_datadata set tofull_data.
The final full_data data frame should have 107 observations and 7 variables.
We will use full_data for the remainder of the assignment. (Note that there are more than 100 observations total due to ties at 50.)
Which states have the most schools in the
full_datadata set? Find the number of schools by state. Then, order these states from greatest to least and return the 5 states with the most schools on the list. Report these 5 states.Which states do not have a school in the
full_datadata set? Use thestate_datadata set and an appropriate join to help answer this question. Return a data set with two variables, state abbreviations and state population, in order from greatest population to least. Show all code and output, and print the state abbreviation and populations. What is the state with the largest population that does not have a school in thefull_datadata set?
Hint: to grab a column name that begins with a number, you need to put the column name in appropriate dashes, e.g.
data %>%
select(`2020pop`)- Recreate the below plot. Use a
dplyrcommand to create the variable. After recreating the plot, discuss what patterns you observe. Hint: create a new variable to determine if a state is in the top 25 or bottom 25 of economic outlook in 2020. Note that 1-25 is the top ranking while 26-50 is the bottom ranking.
- Is there a relationship between state population and the number of schools it has in the
full_datadata set? To answer this question, first use the code from exercise 2 to create a new data set calledcountswith the the number of schools by state as a column.
Then, use a join to add
countsto thefull_datadata set and make a scatter plot with a state’s 2020 population as the x-axis variable and the number of schools as the y-axis variable. Add a line of best fit using themethod = lmoption. Include an informative title and axes labels.Finally, describe what you observe. Is there a positive or negative relationship? Do most points fall near the line?
Let’s now focus on North Carolina schools in the
full_datadata set. For these schools, create a new variable that indicates the change in ranking from 2021 to 2022, where a positive value indicates an increased ranking (e.g., if a school went from 11 to 10, you would want this variable to have a value of positive 1.) Finally, return a tibble that shows the name of the NC schools and the new variable you created. Discuss what you observe.Does the economic outlook and populations of states where national universities are located differ from those where national liberal arts colleges are located?
Using the
full_datadata set, find the mean population and mean economic outlook ranking for national universities vs national liberal arts colleges.Report your results in a 2x3 tibble and discuss.