Advanced ggplot2: Faculty Salary Analysis (Practical-i)

From Basic Layers to Multi-Dimensional Visuals

Author

Abdullah Al Shamim

Published

February 13, 2026

Notebook Introduction (Metadata)

Aim: Extract insights from complex datasets using R and ggplot2 and customize graphs to publication-quality levels.
Dataset: car package’s Salaries dataset (information regarding 9-month salaries of professors in the US).
Key Learnings: Use of geom_jitter(), custom color scales, modifying axis tick marks, and positioning legends inside the plot.

1. Environment Setup and Data Overview

At the beginning, we will load the necessary packages and take a look at the data.

Code

# Installing and loading required packages
# install.packages("car")
library(car)
library(tidyverse)

Code

# Viewing the first few rows of the dataset
# View(Salaries)
head((Salaries), 8)

       rank discipline yrs.since.phd yrs.service  sex salary
1      Prof          B            19          18 Male 139750
2      Prof          B            20          16 Male 173200
3  AsstProf          B             4           3 Male  79750
4      Prof          B            45          39 Male 115000
5      Prof          B            40          41 Male 141500
6 AssocProf          B             6           6 Male  97000
7      Prof          B            30          23 Male 175000
8      Prof          B            45          45 Male 147765

2. Basic Scatter Plot

Let’s look at the relationship between years passed since PhD (yrs.since.phd) and salary (salary).

Code

# Step 1: Basic Scatter Plot
Salaries %>% 
  ggplot(aes(yrs.since.phd, 
             salary, 
             color = rank)) +
  geom_point(size=4, alpha=0.8)

Observation: Here we can see how salary changes according to professional rank. However, many data points are overlapping each other (Overlapping).

3. Customization and Trend Line (Jitter & Smooth)

To reduce data overlapping, we will use geom_jitter and add a regression line.

Code

# Step 2: Advanced Customization
Salaries %>% 
  ggplot(aes(yrs.since.phd, salary)) +
  # geom_jitter spreads the data slightly so that overlapping is reduced
  geom_jitter(aes(color = rank, 
                  shape = discipline)) +
  # Adding a regression line
  geom_smooth(method = lm, 
               se = FALSE,
               color = "red") +
  # Setting colors using preferred hex codes
  scale_color_manual(values = c(
    "AsstProf" = "#A88EF2",
    "AssocProf" = "#FFB222",
    "Prof"     = "#D5A6E2"
  )) +
  theme_test() +
  labs(
    title = "Salary vs. Years since PhD",
    x = "Years since PhD",
    y = "Income",
    Color = "Position",
    Shape = "Research Area"
  )

Code

# Step 3: Advanced Customization
Salaries %>% 
  ggplot(aes(yrs.since.phd, salary)) +
  # geom_jitter spreads the data slightly so that overlapping is reduced
  geom_jitter(aes(color = rank, 
                  shape = discipline), size=4) +
  # Adding a regression line
  geom_smooth(method = lm, 
               se = FALSE,
               color = "red")

Code

# Step 4: Advanced Customization
Salaries %>% 
  ggplot(aes(yrs.since.phd, salary)) +
  # geom_jitter spreads the data slightly so that overlapping is reduced
  geom_jitter(aes(color = rank, 
                  shape = discipline), size=4) +
  # Adding a regression line
  geom_smooth(method = lm, 
               se = FALSE,
               color = "red") +
  # Setting colors using preferred hex codes
  scale_color_manual(values = c(
    "AsstProf" = "#A88EF2",
    "AssocProf" = "#FFB222",
    "Prof"     = "#D5A6E2"
  ))

Code

# Step 5: Advanced Customization
Salaries %>% 
  ggplot(aes(yrs.since.phd, salary)) +
  # geom_jitter spreads the data slightly so that overlapping is reduced
  geom_jitter(aes(color = rank, 
                  shape = discipline), size=4) +
  # Adding a regression line
  geom_smooth(method = lm, 
               se = FALSE,
               color = "red") +
  # Setting colors using preferred hex codes
  scale_color_manual(values = c(
    "AsstProf" = "#A88EF2",
    "AssocProf" = "#FFB222",
    "Prof"     = "#D5A6E2"
  )) +
  theme_test() +
  labs(
    title = "Salary vs. Years since PhD",
    x = "Years since PhD",
    y = "Income",
    Color = "Position",
    Shape = "Research Area"
  )

4. Intermediate Customization (Boxplots & Axis Styling)

Now we will see the salary distribution according to rank and make the axis labels more readable.

Code

# Step 1: Intermediate Customization
Salaries %>%
  filter(salary < 150000) %>% # Filtering high-salary outliers
  ggplot(aes(rank, salary, fill = sex)) + # fill = sex added here for insights
  geom_boxplot() +
  scale_fill_manual(values = c(
    "Female" = "#A88EF2",
    "Male"   = "#D5A6E2"
  )) +
  theme_test() +
labs(
    title = "Faculty Salary by Rank and Gender",
    x = "Academic Rank",
    y = "Annual Salary",
    fill = "Gender"
  )

Code

# Step 2: Intermediate Customization
Salaries %>%
  filter(salary < 150000) %>% # Filtering high-salary outliers
  ggplot(aes(rank, salary, fill = sex)) + # fill = sex added here for insights
  geom_boxplot() +
  scale_fill_manual(values = c(
    "Female" = "#A88EF2",
    "Male"   = "#D5A6E2"
  )) +
  theme_test() +
  labs(
    title = "Faculty Salary by Rank and Gender",
    x = "Academic Rank",
    y = "Annual Salary",
    fill = "Gender"
  ) +
  
  # Beautifying axis labels using 'New Line' (\n)
  scale_x_discrete(breaks = c("AsstProf", "AssocProf", "Prof"),
                   labels = c("Assistant\nProfessor", 
                              "Associate\nProfessor", 
                              "Professor"))

Code

# Step 3: Intermediate Customization
Salaries %>%
  filter(salary < 150000) %>% # Filtering high-salary outliers
  ggplot(aes(rank, salary, fill = sex)) + # fill = sex added here for insights
  geom_boxplot() +
  scale_fill_manual(values = c(
    "Female" = "#A88EF2",
    "Male"   = "#D5A6E2"
  )) +
  theme_test() +
  labs(
    title = "Faculty Salary by Rank and Gender",
    x = "Academic Rank",
    y = "Annual Salary",
    fill = "Gender"
  ) +
  
  # Beautifying axis labels using 'New Line' (\n)
  scale_x_discrete(breaks = c("AsstProf", "AssocProf", "Prof"),
                   labels = c("Assistant\nProfessor", 
                              "Associate\nProfessor", 
                              "Professor")) +
  
  # Setting Y-axis labels in currency format
  scale_y_continuous(breaks = c(50000, 100000, 150000, 200000),
                     labels = c("$50k", "$100k", "$150k", "$200k"))

Code

# Step 4: Intermediate Customization
Salaries %>%
  filter(salary < 150000) %>% # Filtering high-salary outliers
  ggplot(aes(rank, salary, fill = sex)) + # fill = sex added here for insights
  geom_boxplot() +
  scale_fill_manual(values = c(
    "Female" = "#A88EF2",
    "Male"   = "#D5A6E2"
  )) +
  theme_test() +
  labs(
    title = "Faculty Salary by Rank and Gender",
    x = "Academic Rank",
    y = "Annual Salary",
    fill = "Gender"
  ) +
  
  # Beautifying axis labels using 'New Line' (\n)
  scale_x_discrete(breaks = c("AsstProf", "AssocProf", "Prof"),
                   labels = c("Assistant\nProfessor", 
                              "Associate\nProfessor", 
                              "Professor")) +
  
  # Setting Y-axis labels in currency format
  scale_y_continuous(breaks = c(50000, 100000, 150000, 200000),
                     labels = c("$50k", "$100k", "$150k", "$200k")) +
  
  # Moving the legend to the empty space inside the graph
  theme(legend.position = c(0.15, 0.80))

Pro-tips and Shortcuts:

geom_jitter(): When you have a lot of data points, it is mandatory to use this. It gives a visual clue as to where the data density is highest.
scale_x_discrete(): Using \n to shorten long labels (Assistant Professor) is a great practice.
Legend Position: The value of legend.position = c(x, y) stays between 0 and 1. Using this saves space around the graph.

Great work! You have successfully transformed a complex salary dataset into a professional graph.