Not sure which lesson to do first? Find your level and get a plan →

Math · Statistics & Probability · Grade 6-8

Variability

Variability measures how spread out or bunched-up a data set is around its center.


Quick Answer

Variability measures how spread out or bunched-up a data set is around its center. Use it whenever you compare groups that might share an average but differ in how consistent they are: two players' scores, two bus routes' travel times, the weights of loaves from two bakeries. Low variability means the values sit close together and are easy to predict; high variability means they swing widely, so any single value could land far from the average. Range, IQR and standard deviation are the numbers that measure it.

Why This Matters

An average alone can make a risky choice look safe. A bus that averages 20 minutes but sometimes takes 45 will make you late; one that always takes 19 to 21 will not. Knowing variability lets you judge reliability, spot an inconsistent player on a team, and decide how far to trust any prediction built on the mean.

Intuitive Explanation

Picture two dartboards after five throws each. On the first, all five darts sit in a tight cluster around the bullseye. On the second, the darts land at the top edge, the bottom edge, the far left, the far right and one near the middle. Average the positions and both boards come out centered on the bullseye. What separates them is how far each dart lands from that center, and that is variability.

Numbers work the same way. The scores 70, 72, 71, 69 have a mean of 70.5, and every score is within 1.5 points of it. The scores 50, 100, 60, 95 have a mean of 76.25, but the values sit up to about 26 points away from it. In the first set any one score is a good guess for the others; in the second, knowing one score tells you little about the next.

Variability is not the same as a big average. The heights 180, 181 and 182 cm are large numbers with almost no variability, while 1, 5 and 9 are small numbers that are spread out. Ask how far apart the values are, not how big they are. Variability is also the broad idea; range and standard deviation are two different ways of putting a number on it.

Think of it like this. Variability is the reliability of a bus route. Two routes can both average 20 minutes, but the one that takes anywhere from 10 to 40 minutes makes you leave early every day, while the steady one lets you plan to the minute. The picture covers spread around a typical value; it does not say which measure, range or standard deviation, to compute.

Core idea. Variability is the degree to which data points differ from each other and from the center.


When to Use

Reach for variability when a question asks how consistent, reliable or spread out values are, or when two groups have similar averages and you still need to tell them apart. Words like varies and predictable point here too. If the question asks for a typical or usual value, you want a center (mean or median) instead; if it is about one value far from all the others, check whether it is an outlier.

Tip. Ask: Am I describing how scattered the values are, separate from where they center?

How to Recognize It

Before you talk about variability, check whether the question is about how far the values sit from each other or about where they sit on the number line.

  1. Do the groups being compared have about the same average?

    If yes, any difference between them lies in how spread out they are, so variability is the comparison to make. If the averages differ a lot and that is what the question is about, compare centers with the mean or median.

  2. Would adding the same number to every value leave the answer unchanged?

    If yes, you are dealing with spread: raising every score by 5 points moves the center but keeps every gap the same. If the answer would change, as an average does, the question is about center.

  3. Does the question want one specific number, such as the range or the IQR, rather than a description?

    If yes, compute that measure: the range from the extremes, the IQR from the quartiles. Variability is the general idea those numbers measure, so a claim like 'route B is less consistent' needs one of them as evidence.

  4. Is the question about telling a random wobble apart from a real change?

    Then you are in noise or signal-vs-noise territory. Variability counts all of the spread, whatever causes it; noise is only the part with no explanation.

Variability compared with Center (mean/median), Range (a statistic), and Noise

Center, range, noise and variability all come from one list of values, so a claim like 'scores varied a lot' can be mistaken for any of them. Ask whether it is about the typical value, the distance between the extremes, all of the scatter, or only its unexplained part.

Variability

Meaning
Use this when you are comparing how consistent or scattered data is, not where it is centered.
Key test
Am I describing how scattered the values are, separate from where they center?
Example
Player A scores 10,12,11,12,10; Player B scores 2,20,4,18,11. Both average 11 — which is more variable?

Center (mean/median)

Meaning
Tells where data clusters, not how scattered it is.
Key test
Use when you want the typical value, not the spread.
Formula
xˉ=∑xn
Example
Average score of 80

Range (a statistic)

Meaning
Is one specific number for spread, while variability is the broader idea.
Key test
Use when you want a single quick spread number from the extremes.
Formula
max−min
Example
Spread of 96 from 2 to 98

Noise

Meaning
Is the random, unexplained part of variation specifically, not all variability.
Key test
Use when separating random fluctuation from real pattern.
Example
Static around the true signal

Formula & Notation

How to read it: σ (sigma) denotes population standard deviation, s denotes sample standard deviation, and σ2 or s2 denote variance. R often denotes range.

Worked Examples

Example 1 — Comparing two players

Easy

Player A scores 10,12,11,12,10; Player B scores 2,20,4,18,11. Both average 11 — which is more variable?
Show the steps and the answer
  1. Confirm the centers match: 10+12+11+12+10=55 and 2+20+4+18+11=55, so both means are 55÷5=11.

    With equal means, any difference between the players has to come from spread, not from where they center.

  2. Find how far each of A's scores is from 11: 1,1,0,1,1.

    Distances from the mean show how tightly a player's games cluster around their average.

  3. Do the same for B: 9,9,7,7,0.

    B's games land as far as 9 points from the average, on both sides.

  4. Compare ranges: A runs from 10 to 12, a range of 2; B runs from 2 to 20, a range of 18.

    The range puts one number on each player's spread, and B's is nine times A's.

Answer: Player B is more variable

Takeaway: Equal means can hide very different variability: A scores close to 11 every game, while B could score 2 or 20.

Example 2 — A center question

Standard

For Player A's scores 10,12,11,12,10, what is the typical score?
Show the steps and the answer
  1. Read what is asked: a typical score, one value that stands for the whole set.

    'Typical' asks where the data sits, which is a center question, even though these scores appeared in a spread problem a moment ago.

  2. Add and divide: (10+12+11+12+10)÷5=55÷5=11.

    The mean is the center measure that uses every score.

  3. Cross-check with the median: sorted, the scores are 10,10,11,12,12, and the middle one is 11.

    Mean and median agree because A's scores are balanced around 11, so 11 is a fair answer either way.

Answer: Mean =11

Takeaway: A question about the typical value needs a center. Spread would say how far A's games stray from 11, which this question did not ask.

Example 3 — Find the mistake: equal means, equal loaves?

Application

Two bakeries sell loaves meant to weigh 500 g. Bakery P's loaves weigh 498, 502, 500, 499 and 501 g; Bakery Q's weigh 470, 530, 515, 485 and 500 g. Lee writes: "Both bakeries average 500 g, so their loaves are equally consistent." Find the mistake and correct it.
Show the steps and the answer
  1. Check Lee's means: each set totals 2500 g, and 2500÷5=500 g.

    Lee's arithmetic is right; the error is in what he concludes from it.

  2. Find each range: P is 502−498=4 g; Q is 530−470=60 g.

    Consistency is a question about spread, and a mean carries no information about spread.

  3. Compare the worst loaves: Q sells loaves 30 g light and 30 g heavy, while P's are never more than 2 g off.

    Q's mean is 500 g only because its heavy and light loaves cancel out when averaged.

Answer: Bakery P is far more consistent: its range is 4 g, Bakery Q's is 60 g, even though both average 500 g.

Takeaway: Equal means do not mean equal data. Report a spread next to every mean before calling two groups alike.

Where You'll See It

  • Choosing a bus route to school

    Route 1 takes 18, 20, 19, 21 and 22 minutes over a week; Route 2 takes 10, 35, 12, 30 and 13. Both average 20 minutes, but Route 2's range is 25 minutes against Route 1's 4. To be on time by Route 2 you must allow for its 35-minute days, 15 minutes more than the average, so the steadier route is the better choice.

  • Filling bottles in a juice factory

    A machine fills bottles labelled 500 mL. If fills run from 495 to 505 mL, every bottle is close to its label. If they run from 470 to 530 mL, the average can still be 500 mL, yet some customers get 30 mL less than they paid for. Factories track the spread of fills, not just the mean, to catch a machine that needs adjusting.

  • Packing for two cities

    San Francisco and Kansas City both average roughly 13 to 15 °C over a year. San Francisco's monthly averages stay between about 11 and 18 °C, while Kansas City's run from about -1 °C in January to about 27 °C in July. The yearly average is nearly the same; the variability decides whether you need a winter coat.

Common Mistakes

  1. The slip: Reporting only the mean and stopping

    Instead: a mean with no spread hides whether the data is consistent.

  2. The slip: Assuming equal means imply equal data

    Instead: same center can come with very different variability.

  3. The slip: Confusing high variability with a high average

    Instead: spread measures distances from the center, not the center's size.


Mini Practice

Try these on your own. Tap Reveal when you want to check.

  1. In 5 games, Ana scored 8,10,9,11,12 points and Bea scored 2,18,5,15,10. Both average 10. Find each range. Who is more consistent?

    Hint: Range is the largest score minus the smallest score.

  2. Which set has more variability: 20,20,20,20 or 18,19,21,22? Give the range of each.

    Hint: Do all the values in the first set match?

  3. Find the mean absolute deviation of 4,6,8,10,12. The mean is 8.

    Hint: Find how far each value is from 8, ignoring signs, then average those distances.

  4. Morning temperatures in degrees Celsius for five days: City X 15,16,15,17,16; City Y 5,25,10,22,18. Which city's weather varies more?

    Hint: Find each city's range and compare.

  5. A principal asks which class scored higher on average. A coach asks whose lap times are more predictable. Which question is about variability?

    Hint: One asks where the values sit; the other asks how scattered they are.

  6. Set A is 100,101,99,100. Set B is 2,9,15,6. Kim says Set A has more variability because its numbers are bigger. Correct her.

    Hint: Variability is about the gaps between the values, not their size.

Want the full set?

50 practice questions for this concept — free to try, every one with a complete worked solution showing the why, not just the answer.

Frequently Asked Questions

Can variability ever be zero?

Yes, when every value is the same. If five friends each have exactly 3 pets, the range is 0, the IQR is 0 and the standard deviation is 0: there is no spread at all. Real measured data almost never does this, because tiny differences show up when you measure carefully, so a spread of exactly zero in real measurements is a hint to check how the data was recorded.

Can two measures of spread disagree about which group varies more?

Yes. Group A, 1, 5, 5, 5, 5, 20, has a range of 19 but an IQR of 0, because its middle values are all 5. Group B, 1, 2, 4, 6, 8, 10, has a range of only 9 but an IQR of 6. The range sees A's two extreme values, while the IQR sees that B's middle is far more scattered. When the measures disagree, say which kind of spread you mean.

Why doesn't a bigger sample make the data less spread out?

Because spread belongs to the thing being measured, not to how many times you measure it. Heights of 12-year-olds vary by many centimeters, and measuring 1,000 of them instead of 20 shows that same spread more clearly. What a bigger sample does shrink is the wobble in the average: the mean of 1,000 heights moves much less from sample to sample than the mean of 20.

Learning Path

← Before

Data (Abstract)
Variability

You are here

You need Data (Abstract) first: a set of values collected about something. Variability adds a second question to ask of any data set, after where it sits: how much do the values differ? Next, Range (Statistics) gives the simplest number for it, largest minus smallest, and Standard Deviation measures the typical distance of values from the mean.

See Also