Simplifying Common Statistical Terms the Lean Way

Descriptive statistics are a goldmine for process improvement. They help you move from “I think there’s a problem” to “Here’s exactly what’s wrong and how often it happens.”

Hey friend,

A few months ago, I made a post about data literacy—how it’s no longer a luxury but a necessity. Some of you took my advice, and I hope you're already seeing the benefits. Others? Still acting like data is a problem for the IT team. My dear, before Excel charts start looking like hieroglyphics, now is the time to get comfortable with data.

This week, we’re diving into statistics—the key to truly understanding data beyond just creating or reading colourful Power BI dashboards.

This post was inspired by a video I stumbled on (shoutout to whoever made it—couldn’t find the link again). The guy in the video called out people who claim to be “data analysts” just because they know how to use a few tools. His question?

"Do you actually understand data? Do you even know what mean, median, or mode is?"

And honestly? He has a point. Tools can make things look fancy, but real insight comes from knowing what your numbers are actually saying.

So today, I’m breaking down some common statistical terms—the Lean way. No jargon, no headaches. Just simple, practical understanding so you can actually use stats to drive improvements in your work without “sweating it”.

---

What is Statistics & Why Does It Matter in Improvement?

Statistics is just a fancy way of saying "making sense of numbers". It helps us spot trends, identify patterns, and make data-driven decisions instead of relying on gut feelings.

In process improvement, statistics help us answer questions like:
- Are things getting better or worse?
- What’s the typical performance?
- How much variation exists in our processes?

Without data and statistics, improvement work would be “vibes and Insha’Allah”. And trust me, that’s not a strategy.

Now, let’s break down some key statistical terms the Lean way.

---

Measures of Central Tendency (Where Your Data "Centers" Around)

1. Mean (Average)

Textbook Version: The sum of all values divided by the total number of values.
What does that even mean?
If your average customer wait time is 20 minutes, but some people are waiting 5 minutes while others are waiting 45 minutes, is that “average” really good for business? Mean gives a general idea, but it won’t show extreme values.

2. Median (Middle Value)

Textbook Version: The middle number when all values are arranged in order.
What does that even mean?
If you’re analysing customer service response times and the median is 12 minutes, it means half of your customers are getting served in 12 minutes or less. Unlike the mean, the median does not consider extreme values (e.g., that one unlucky person who waited 3 hours).

3. Mode (Most Common Value)

Textbook Version: The value that appears most frequently in a dataset.
What does that even mean?
If most customers complain about slow delivery times on Fridays, you’ve just found a pattern worth investigating. The mode helps identify common occurrences in your data.

---

Measures of Variability (How Scattered Your Data Is)

4. Variance (How Far Data Points Are from the Mean)

Textbook Version: The average of the squared differences from the mean.
What does that even mean?
If you track daily sales and some days you sell 5 units while other days you sell 500 units, your variance is high. That level of unpredictability will make it hard to plan inventory and staffing properly.

5. Standard Deviation (The "Spread" of Your Data)

Textbook Version: The square root of variance, showing how much data deviates from the mean.
What does that even mean?
If your manufacturing process has a low standard deviation, it means your products are consistently good. If it’s high, some products are amazing, some are absolute disasters—and that’s not great.

6. Range (Difference Between the Highest & Lowest Value)

Textbook Version: The highest value minus the lowest value.
What does that even mean?
If your customer complaint resolution time ranges from 5 minutes to 3 days, something isn’t right. Range helps spot extreme differences in performance.

---

Measure of Frequency Distribution (How Often Something Happens)

7. Count (Number of Occurrences)

Textbook Version: The total number of times an event happens.
What does that even mean?
If customer complaints about delivery delays happen 50 times in a month, that’s a huge red flag. Counting events helps prioritise issues.

---

Final Thoughts

Descriptive statistics are a goldmine for process improvement. They help you move from “I think there’s a problem” to “Here’s exactly what’s wrong and how often it happens.”

Some businesses have Business Intelligence (BI) teams to do this heavy lifting. Others? You’re on your own. Either way, the more you understand data, the better equipped you’ll be to drive improvements.

So, let me ask you:
Do you really understand your data, or do you just find colourful charts interesting?

Until next time, keep improving!

– Tomiwa
Lean Process Improvement Enthusiast
(Still trying to figure out the variance in my bank alerts.)