R Notebook

IBM HR Analytics Employee Attrition & Performance - EDA

Objective:

Find out leading factors contributing to employee attrition

Approach:

Use of R programming to analyse the data set

View dataset

head(df)

It look like the dataset have categorical value which is encoded into numeric . Here are labels: Education 1 ‘Below College’ 2 ‘College’ 3 ‘Bachelor’ 4 ‘Master’ 5 ‘Doctor’

EnvironmentSatisfaction 1 ‘Low’ 2 ‘Medium’ 3 ‘High’ 4 ‘Very High’

JobInvolvement 1 ‘Low’ 2 ‘Medium’ 3 ‘High’ 4 ‘Very High’

JobSatisfaction 1 ‘Low’ 2 ‘Medium’ 3 ‘High’ 4 ‘Very High’

PerformanceRating 1 ‘Low’ 2 ‘Good’ 3 ‘Excellent’ 4 ‘Outstanding’

RelationshipSatisfaction 1 ‘Low’ 2 ‘Medium’ 3 ‘High’ 4 ‘Very High’

WorkLifeBalance 1 ‘Bad’ 2 ‘Good’ 3 ‘Better’ 4 ‘Best’

General Overview:

distribution of employees across different age groups?

library(ggplot2)

# Create a histogram
ggplot(df, aes(x = Age)) +
  geom_histogram(binwidth = 5, fill = "skyblue", color = "black", alpha = 0.7) +
  labs(title = "Distribution of Employees Across Age Groups",
       x = "Age",
       y = "Number of Employees") +
  theme_minimal()

How many employees have experienced attrition?

# Using sum() function
attrition_count <- sum(df$Attrition == "Yes")
print(paste("Number of employees who have experienced attrition: ", attrition_count))

[1] "Number of employees who have experienced attrition:  237"

Demographics:

What is the distribution of gender and marital status among employees?

ggplot(df, aes(x = MaritalStatus, fill = Gender)) +
  geom_bar(position = "dodge", color = "white") +
  labs(title = "Distribution of Gender and Marital Status Among Employees",
       x = "Marital Status",
       y = "Number of Employees",
       fill = "Gender") +
  theme_minimal()

How is attrition related to gender and marital status?

ggplot(df, aes(x = Attrition, fill = Gender)) +
  geom_bar(position = "dodge", color = "white") +
  labs(title = "Relation of Attrition and Gender Among Employees",
       x = "Attrition",
       y = "Number of Employees",
       fill = "Gender") +
  theme_minimal()

ggplot(df, aes(x = Attrition , fill = MaritalStatus))+ geom_bar(position = "dodge", color = "white")+labs(title = "Relation of Attrition and Maritial Status Among Employees",
       x = "Attrition",
       y = "Number of Employees",
       fill = "Gender") +
  theme_minimal()

Work Environment:

What is the distribution of business travel frequency?

ggplot(df, aes(x = BusinessTravel, fill = BusinessTravel)) +
  geom_bar() +
  labs(title = "Distribution of Business Travel Frequency",
       x = "Business Travel Frequency",
       y = "Number of Employees") +
  theme_minimal()

How does the departmental distribution look?

ggplot(df, aes(x = Department, fill = Department)) +
  geom_bar() +
  labs(title = "Departmental Distribution",
       x = "Department",
       y = "Number of Employees") +
  theme_minimal()

What is the average daily rate for employees?

average_daily_rate <- mean(df$DailyRate)
print(paste("Average Daily Rate for Employees: $", round(average_daily_rate, 2)))

[1] "Average Daily Rate for Employees: $ 802.49"

Job-related Factors:

How does job satisfaction vary across different roles and levels?

ggplot(df,aes(x=JobRole , y= JobLevel))+geom_line() + labs(title = "Distribution of Job Roles and Level" , x= "Job Role " , y = "Job Level") +theme_minimal()+theme(axis.text.x = element_text(angle = 90, hjust = 1))

What is the distribution of job roles and levels?

ggplot(df, aes(x = JobRole, fill = factor(JobLevel))) +
  geom_bar(position = "stack") +
  labs(title = "Distribution of Job Roles and Levels",
       x = "Job Role",
       y = "Number of Employees",
       fill = "Job Level") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 90, hjust = 1))  

Are there any patterns in the relationship between job involvement and attrition?

ggplot(df, aes(x = JobInvolvement, fill = Attrition)) +
  geom_bar(position = "stack") +
  labs(title = "Relationship Between Job Involvement and Attrition",
       x = "Job Involvement",
       y = "Number of Employees",
       fill = "Attrition") +
  theme_minimal()

How does attrition relate to monthly income and other compensation-related factors?

ggplot(df, aes(x = Attrition, y = MonthlyIncome, fill = Attrition)) +
  geom_boxplot() +
  labs(title = "Monthly Income by Attrition",
       x = "Attrition",
       y = "Monthly Income",
       fill = "Attrition") +
  theme_minimal()

NA
NA

Employee Satisfaction:

How satisfied are employees with their work environment, relationships, and work-life balance?

ggplot(df, aes(x = factor(Attrition), fill = factor(Attrition))) +
  geom_bar(aes(y = EnvironmentSatisfaction), position = "dodge", stat = "summary", fun = "mean") +
  geom_bar(aes(y = RelationshipSatisfaction), position = "dodge", stat = "summary", fun = "mean") +
  geom_bar(aes(y = WorkLifeBalance), position = "dodge", stat = "summary", fun = "mean") +
  labs(title = "Satisfaction with Work Environment, Relationships, and Work-Life Balance",
       x = "Attrition",
       y = "Mean Satisfaction Score",
       fill = "Attrition") +
  scale_fill_manual(values = c("No" = "lightblue", "Yes" = "lightcoral")) +
  theme_minimal()

Does satisfaction with the work environment correlate with attrition?

ggplot(df, aes(x = Attrition, y = EnvironmentSatisfaction, fill = Attrition)) +
  geom_boxplot() +
  labs(title = "Environment Satisfaction by Attrition",
       x = "Attrition",
       y = "Environment Satisfaction",
       fill = "Attrition") +
  theme_minimal()

Performance Metrics:

What is the distribution of performance ratings?

ggplot(df, aes(x = PerformanceRating, fill = factor(PerformanceRating))) +
  geom_bar() +
  labs(title = "Distribution of Performance Ratings",
       x = "Performance Rating",
       y = "Number of Employees",
       fill = "Performance Rating") +
  theme_minimal()

How do performance ratings relate to attrition?

t_test_result <- t.test(PerformanceRating ~ Attrition, data = df)
# Display the result
print(t_test_result)

    Welch Two Sample t-test

data:  PerformanceRating by Attrition
t = -0.10999, df = 331.22, p-value = 0.9125
alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
95 percent confidence interval:
 -0.05350780  0.04784086
sample estimates:
 mean in group No mean in group Yes 
         3.153285          3.156118 

Training and Development:

How many training sessions did employees undergo last year?

total_training_sessions <- sum(df$TrainingTimesLastYear)

# Display the result
print(paste("Total Training Sessions Last Year: ", total_training_sessions))

[1] "Total Training Sessions Last Year:  4115"

Is there a correlation between training and attrition?

t_test_result_2 <- t.test(TrainingTimesLastYear ~ Attrition, data = df)
# Display the result
print(t_test_result_2)

    Welch Two Sample t-test

data:  TrainingTimesLastYear by Attrition
t = 2.3305, df = 339.56, p-value = 0.02036
alternative hypothesis: true difference in means between group No and group Yes is not equal to 0
95 percent confidence interval:
 0.03251776 0.38439273
sample estimates:
 mean in group No mean in group Yes 
         2.832928          2.624473 

Temporal Analysis:

Have there been any trends in attrition over the years?

ggplot(df, aes(x = YearsAtCompany, fill = Attrition)) +
  geom_bar(position = "stack") +
  labs(title = "Trends in Attrition Over the Years",
       x = "Years at Company",
       y = "Number of Employees",
       fill = "Attrition") +
  theme_minimal()

How do various factors change with the number of years an employee has spent in the current role or with the current manager?

ggplot(df, aes(x = YearsInCurrentRole, y = MonthlyIncome, color = Attrition)) +
  geom_point() +
  labs(title = "Relationship Between Monthly Income and Years in Current Role",
       x = "Years in Current Role",
       y = "Monthly Income",
       color = "Attrition") +
  theme_minimal()+ geom_jitter()

Overtime:

How many employees work overtime, and is there a correlation with attrition?

employees_overtime <- sum(df$OverTime == "Yes")

# Display the result
print(paste("Number of Employees Working Overtime: ", employees_overtime))

[1] "Number of Employees Working Overtime:  416"

ggplot(df, aes(x = OverTime, fill = Attrition)) +
  geom_bar(position = "fill") +
  labs(title = "Relationship Between Overtime and Attrition",
       x = "Overtime",
       y = "Proportion",
       fill = "Attrition") +
  scale_y_continuous(labels = scales::percent_format(scale = 1)) +  # Display y-axis in percentage
  theme_minimal()

Summary

It look like the leading factors of attrition are salary(income), job satisfaction , lack of work life balance. Increase in training hour could potentially decrease atttrition .