Mean, Median and Spread Without the Panic

Here's a scene I watch play out all the time. A GRE Test Taker is moving nicely through a Quant section, hits the words standard deviation, and freezes. Pencil stops. Eyes go wide. Then they try to remember a formula with a square root and a Greek letter in it. Two minutes later, they guess.

Here's the blunt verdict. That panic is costing you points, and the formula was almost never the point. Most GRE statistics questions don't care whether you can grind through a long calculation. They care whether you know what mean, median and spread actually DESCRIBE. Learn the ideas and these questions get FAST.

And they're not rare. ETS lists basic statistics like mean, median, range and standard deviation in the data analysis part of the Quant section. You get two scored Quant sections, 20 questions in 35 minutes each. That's not much time to spend panicking.

Three ideas, three plain jobs

• Mean is the average. Add up the numbers, then divide by how many numbers there are. Think of it as the balance point of the list. Every number pulls on it, and a number far away pulls HARD.
• Median is the middle number once you put the list in order. With an even count, it's halfway between the two middle numbers. It only cares about position, so one wild number at the end barely moves it.
• Spread is how far the numbers sit from the middle. The range is the biggest number minus the smallest. The standard deviation is, roughly, the typical distance of a number from the mean. Numbers bunched together give a small one. Numbers scattered far apart give a big one.

That's it. Now stop and think. If I change the biggest number in a list from 50 to 5,000, what happens to the median? Don't read another word until you have an answer.

Nothing happens. The middle number is still the middle number. The mean, on the other hand, jumps way up. That one idea answers a surprising number of GRE questions.

Try one: mean versus median

This one is a Quantitative Comparison, a GRE format where you're given two quantities and you decide which one is bigger, whether they're equal, or whether you can't tell.

Notice what you did NOT do. You didn't set up a giant inequality first. You wrote the list in order, tested a couple of numbers and let the results talk. It's the same lesson as our post on ratios and the equations you don't need.

Spread without the formula

The GRE gives you an on-screen calculator, and some people take that as an invitation to compute every standard deviation by hand. THAT is the long way. Two facts do most of the work for you:

• Add the same number to every value, and the mean moves by that number, but the spread stays EXACTLY the same. Every value slid over together, so the distances between them didn't change.
• Multiply every value by the same positive number, and the mean and the standard deviation both get multiplied by it. Triple every value and the spread triples too.

The mistakes I see every week

1. Memorizing the formula and stopping there. You can recite it, but you can't tell which list is more spread out without punching buttons. INSTEAD: for every list, ask "how far are these numbers from the middle?" before you touch the calculator.
2. Finding the median of an unsorted list. The middle of 9, 2, 7 is not 2. INSTEAD: write the list in order on paper, every single time. It takes five seconds.
3. Assuming the mean and median are the same. They're only guaranteed equal in special cases, like evenly spaced lists. INSTEAD: check whether one end has a number far away from the rest. If it does, the mean gets pulled toward it.
4. Treating range as the whole story on spread. INSTEAD: use the range as a first look, then check where the OTHER numbers sit.
5. Testing one value on a Quantitative Comparison and calling it done. INSTEAD: test at least two values, and make one of them sit right near the edge of what's allowed. That's where answer D hides.
6. Doing all of it in your head. You think you're "smart enough" to hold five numbers and their distances in memory. You're not, and neither am I. That's pride, and pride makes silly mistakes. INSTEAD: take notes. Same rule as when you label everything in geometry before you solve anything.

Why this makes you faster

I'll bet the stats questions that scared you most never needed the formula at all. They needed you to see that one number was dragging the mean, or that two lists had the same spread, or that a test value near the edge flips the answer. Those are thinking moves, not math moves. And thinking moves take seconds once you've trained them.

That's the real point. The GRE is a predictable, standardized test. The same few ideas about mean, median and spread show up again and again, dressed in different numbers. You don't need to study MORE statistics. You need to TRAIN the handful of ideas that keep showing up until they're automatic.

So here's your challenge for today. Grab a sheet of paper and write three lists of five numbers. Make one tight, one with a single far-off number and one scattered. For each list, write the mean and the median. Then rank the three by spread WITHOUT calculating anything. Only then check yourself with a calculator. If your ranking was wrong, you are NOT allowed to say "I have no idea why." Find the number that fooled you and write down what it taught you. Then do it again tomorrow with new lists, and when you're ready, find the statistics questions in one of ETS's free POWERPREP practice tests and see how fast they fall.