Why The Great Resignation

The Great Resignation: People Analytics R Project

Imagine your boss just grilled you on the numbers that day and more so lately than ever. It was the busiest time of the year, record-breaking numbers, and you did the best you could with the approved staffing. Bottom line, we needed more staff to handle the workload. In the bathroom, looking in the mirror after being upset, as it's not emotionally intelligent to show others you're human. Lost a lot of weight from the stress of work, breaking out, and losing clumps of hair. Barely see your family and weeks go by so fast it seems like you're getting paid every other day. Working around the clock to ensure the work gets done, so you never go to the doctor or dentist to make sure everything is okay. A specific song comes on the radio, "You're going to miss me when I'm gone" ?? was playing. You look at the reflection in the mirror, stop and realize the job took over you. At that very moment, it was a sign to leave. It was time to find a new opportunity that didn't take so much from you.

Why this Project?

This is why I decided to do this project. I had been with the company for over 6 years so it was difficult to just walk away. My team of 32 was tired as well. They were expected to reach very high quotas with the same quality work and work overtime. I loved being a manager, my staff, I loved the job, but the reason for quitting had nothing to do with any of those things, but everything to do with my manager's decisions. There was no convincing or showing facts to make my point, as she had her mind made up. I finally made the decision, and gave my two weeks notice to my boss so she could find a replacement. I thanked her for giving me the opportunity, and I never looked back.

She was a new manager and so was I. How could she teach me what she didn't know? I was excited about the opportunity to finally become a manager, but I should have considered the reason the old manager left... they passed away. I didn't even think twice about how the job basically killed her and I was next. My manager before this had to leave because she got cancer. A very high stress environment to say the least. There are jobs just like this that exist, maybe not exactly like this, but the reasons for leaving are the same. Unreal expectations, even for me as I always exceeded expectations, and I thrived on being busy. I had just received manager of the quarter, and there was talk about becoming the next director. What happened? How did this script flip so quickly? I learned a lot of very important lessons that helped me become an even better manager when the time presented itself. When I left, I was told that they had to hire at least 3 people to do the work I was doing. At the end, that's all I wanted and asked for was approval for more staff.

So for this project, you ask why do people leave their job? What is the main reason?

What Readers Will Gain

By reading this article, you will learn.

• Twice as many men than women that left, specifically single men that rarely travel for work
• The top three jobs with the most attrition: Sales, Lab, and Research Scientist, departments all require overtime
• The top two departments with the most attrition are Research Development with 133 that left and 92 left in Sales
• 35.9% of employees that left had the current manager for less than 1 year, 21.1% 2 years, 13.1% for 7 years
• 16.1% attrition rate for this specific dataset vs what is considered a good rate 10-15%

Dataset Details

This dataset was found here. This data set was an augmented version created by real IBM data scientists, but isn't exactly 100% real data. It's probably the most commonly used data set for People Analytics.?This data set has 1470 rows where each row is an employee. There are 35 different columns that describe that employee. The most important column is the "Attrition" column which describes whether that employee left or stayed at the company. Age ranges are from 18 to 60, with a median age of 36.

What causes people to resign? Is it because of their boss? The workload, getting passed on promotions, no increase in pay, too much traveling, no work life balance, asking for time off was frowned upon, short staff, no rewards, no team building, a toxic work environment, personal reasons, sickness, or death?

Analysis Process

Let's first define what attrition is exactly versus employee turnover. "Employee turnover is the loss of talent in the workforce over time due to layoffs, terminations, location transfers, resignations, retirements or other separations. Employee turnover should not be mistaken for employee attrition...refers to the natural reduction of staff due to resignation, retirement or personal health issues, with no intention of replacing the position." (Business.com 2024)

For this project, I'm a People Data Analyst?intern for?IBM?in the Human Resources department. Recently, there's been lots of people leaving the company and my boss wants me to explore and find out why. It's easy to just share my story and domain knowledge of potential reasons, but let's show some stats behind it by using R.

Correlation Matrix

The company wants to get an overview of how some of the most important demographics correlate.?Looking through these numbers, the closer to 1 or -1 the values are, the stronger the linear relationship.?I ran the following to find out if any are correlated.

The ones most highly correleted are monthly income and total working years at .77 confirming that total working years were a stronger predictor of monthly income. Second is age and total working years at .68 which confirms that as they age they gain more experience and working years.

Scatterplots in R

To explore any patterns in the data. I decided to run scatter plot matrices based on the correlation relationships. The following scatter plot matrices contain the three attribute pairs with high correlations and a fourth attribute, gender.

Hypothesis Testing, p-values, Statistical Significance

Next to answer the question, could age be a factor in employee attrition? I created a box plot to display the median age of employees as "Let Go or Resigned" in the attrition column since we are not sure the reason for leaving.

I did the hypothesis test with Welch Two Sample t-test. With one sample that left, and one sample that stayed comparing the average ages of these two sample & calculated a p-value.

To do this in R, I created a new variable called yes_age that is the Age column but only the rows that have attrition as "Yes". Then create another variable called no_age that is the Age column, but only the rows that have "No" in attrition.

Because p-value is less than 0.05, there is a statistically significant difference between the two samples. Those who left were actually younger than those who stayed. We can see that in the mean comparison at the bottom. Note that x is the first array we passed in and y is the second array we passed in. This is also confirmed in the confidence interval, since both those numbers listed below are negative, we know that the first array is smaller than the second, with confidence.

Linear Regression

To determine whether an employee's age could predict their salary, I created two different linear regression models to identify which variables were the most effective predictors of an employee’s income.

The p-value is 2.2e16 (e means 10 to the power of); so nearly 0. Since p is less than 0.05, we can say with 95% confidence this model is statistically significant.?

Since p is small, the model explains around 25% of the variance in the monthly income by just using the age variable.?

In Model 2, I incorporated two variables: age and the total number of years employees have been working, to predict monthly income. The results showed that 59.8% of the variation in income could be explained by these two variables.

Main Takeaways

• Ageism was not a reason for termination. Those who left were younger than those who stayed.
• Doing more exploratory analysis in Tableau deep-diving to get more information I found a specific trend of those that left: Male age 25 (total of 18), age 28 (total of 13), and age 32 (total of 16) in the Research and Development department. As far as Females leaving, in both Sales and Research and Development also the same age 28 (10 total in sales and 11 total in research and development), 19 total had their current manager for less than a year. 34 people that left are exactly the same age and 42 of those that left worked overtime. (I'll share this Tableau viz shortly)
• The columns that correlate the most are: MonthlyIncome-TotalWorkingYears: 0.7728, Age-TotalWorkingYears: 0.6803, and Age-MonthlyIncome: 0.4978
• The model can explains around 25% of the variance in the MonthlyIncome by just using the age variable.?
• The model can explains around 59% of the variance in the MonthlyIncome by just using the employee variable.?
• Recommendations: Investigating the sales and research and development department further to see if it's a specific manager or managers that is the cause for the attrition rate increase. Check for major changes in standard operating procedures, or handling the workload made had a sudden change, or if there is an issue with workload, tools or training not available to succeed or meet quotas, since there is a need for overtime. Check for performance improvement plans to see if they were warranted and scored accurately. Have a conversation with the two managers to see why they think specific age groups might be leaving or let go to get insight on why there seems to be a significant pattern with age. Last, check to see if the manager hired was qualified to be a manager and had management experience prior to becoming the new manager. Sit in on team meetings to see what is being discussed. Provide an anonymous survey to the specific teams encouraging them that what they say will remain anonymous and enforce a retaliation protection policy.

Conclusion and Personal Reflection

In my situation, it was a very difficult decision to leave as I had invested years with the company and wanted to be a director, but this opened a new door for me. I decided at that moment I really needed to spend time with my daughter. She was little, she was not going to be for very long, and I didn't want to miss it. I was grateful that I found work at the best company to work for and it was 100% remote. I loved it. I didn't have to spend money on daycare anymore. I was home when my daughter got out of school. This was the best job a parent could have. I felt like I hit the jackpot. Had I not left the company, who knows what could have happened. This project was relateable as I was almost exactly the same age as the majority of those that left, my manager was a new manager, and I worked crazy overtime to get the work done (I was paid salary so there was no extra pay).

Call to Action

I enjoyed completing this project. Sometimes in life you get caught up in your work, taking care of your family, and you forget the most important thing, which is to take care of yourself. If you are not healthy, who will take care of your job, and who will take care of your family? Would it all be worth it if your last days were spent at work, and not with your family? It's important to get rest, to go to your and your family's doctor appointments, be present at important events, do the work you need to do, go home, and leave work at the door. Prioritize your important work at work to meet deadlines, so that you have a good balance. Make sure you choose work that you love to do or at least really enjoy, and don't just do it for the money, title, or the company name. Do your research to make sure you will fit in with the culture, the values, and the people that work there. This will ensure a happy marriage/relationship, happy kids, and a more fulfilled and enjoyable work life.

I would love to hear your thoughts! Connect with me on LinkedIn, or if you or someone you know is looking to hire a data analyst, let’s talk! Feel free to leave a comment with your thoughts or questions.