Warning: This blog contains extra nerdy feelings about data. Do not read while driving — it could make you drowsy!
In grade school, I was taught the basics of mean, median, and mode. As the years went by, my favorite of the three — and the one I used the most — became mean, or the arithmetic average. The average class grade, the average SAT score, earned run averages, the average height and weight of my kid… mean became the baseline for how I judged data sets.
And don’t get me wrong — mean is still helpful in gauging general direction when trying to answer questions. But since those early years, I’ve discovered that just a handful of outliers (very large or very small numbers) can dramatically impact the average and skew your perspective.
Enter the lesser-known but equally important heroes: median and mode.
You might remember the basic definitions of median and mode, but here’s a quick refresher:
- Median: the middle value in a data set when the numbers are arranged in order.
- Mode: the value that appears most often in a data set.
But let’s look at this in a real-world way.
Let’s pretend I’m in charge of determining salaries for new lab managers. I work in HR, so I research salaries listed on job postings from similar facilities. (In reality, there would be many more formal ways to do this, but again, this is just for demonstrative purposes.)
Here’s what I find:
- $18,000
- $72,000
- $75,000
- $78,000
- $80,000
That $18,000 salary seems like it could be an error, but it might represent a part-time interim manager, someone working very limited hours, or a temporary contract role.
If I only rely on the mean, then the average salary I’d be willing to offer would be $64,600. Now, $64.6K might not seem too far off from the other salaries on paper, but as we all know, that’s a pretty big difference in our bank accounts.
Enter median.
The median salary in this set is $75K. Using this data point along with the mean — and factoring in variables such as organizational budget, benefits, and other compensation factors — I might determine that a more competitive salary would be closer to $73K.
Using two measures of central tendency provides a more accurate picture.
Now let’s talk stain intensity.
In histology labs, stain intensity is often recorded using ordered categories such as:
- 0 = no stain
- 1 = weak
- 2 = moderate
- 3 = strong
These numbers represent rankings rather than exact measurements.
Imagine I ranked 15 stained slides and got the following results:
1, 2, 2, 3, 2, 1, 2, 3, 2, 2, 1, 2, 3, 2, 1
If I calculate the mean intensity, it would be:
(1+2+2+3+2+1+2+3+2+2+1+2+3+2+1) / 15 ≈ 1.93
The result is 1.93, which does not correspond to an actual staining category observed under the microscope. You can’t really classify a slide as having “1.93 staining.”
So how can we determine the staining intensity that occurs most often?
Mode!
In this case, the mode is 2, meaning “moderate staining” appeared more often than any other category.
The mode is often better than the mean for this type of data because stain categories are ordinal, not continuous numerical values. Calculating a mean can produce values like 1.93, which don’t represent an actual staining category seen on a slide. The mode, however, gives a direct and biologically meaningful summary that matches how histology observations are actually recorded and interpreted.
So while mean may still be the statistical celebrity most of us grew up with, median and mode deserve a little more attention. Sometimes using more than one measure of central tendency gives us a clearer, more realistic picture — whether we’re talking salaries, baseball stats, or stained tissue slides.
In the end, the takeaway is simple: no single number can tell the whole story. Mean, median, and mode each bring their own strengths to the table and using them together helps us move beyond oversimplified conclusions and toward more thoughtful, informed decisions. If this kind of “extra nerdy” dive into data has sparked your interest, be sure to join us for Statistical Foundations for Histologists on May 21st at 1 PM ET, where we’ll continue building practical, approachable ways to understand and apply statistics in the lab. This webinar is FREE and exclusively for NSH members. Members can register here.
Written by Connie Wildeman, Director of Education, National Society for Histotechnology
#2026#Blog#Education