Outlier Effects on Center — I can interpret the effect of outliers on mean and median and justify which measure is better. - Math 1

SECTION 1

Historical Context & Motivation

Humans have been summarizing data for centuries. Ancient civilizations recorded harvests and trade goods, but the idea of calculating a single number to represent an entire collection of values is surprisingly modern. The concept of the arithmetic mean (what most people call the "average") became a formal mathematical tool during the Scientific Revolution, while the median gained recognition later as statisticians realized that extreme values could warp averages in misleading ways.

1669 • Christiaan Huygens and Expected Value

Dutch mathematician Christiaan Huygens formalized the idea of an expected (average) value in probability, laying the groundwork for using means to summarize data.

1774 • Laplace and the Median

Pierre-Simon Laplace introduced the median as the value that minimizes absolute deviations, providing an alternative center that resists extreme observations.

1887 • Francis Galton Popularizes Median

Galton advocated using the median when reporting data on human characteristics, noting that a few unusually tall or short individuals could shift the mean and misrepresent the typical person.

1960s • Robust Statistics Emerges

Statisticians like John Tukey developed robust methods—including box plots and trimmed means—specifically designed to handle outliers without being distorted by them.

The central question this lesson addresses is straightforward yet powerful: when your data set contains one or more extreme values—called outliers—which measure of center, the mean or the median, gives a more accurate picture of the 'typical' value? Understanding this distinction is essential for interpreting reports about income, test scores, housing prices, and nearly any real-world data set.

SECTION 2

Core Principles & Definitions

Before we can explore how outliers pull or push summary statistics, we need clear definitions of the key terms involved. Each concept below builds on ideas you already know from Algebra 1 and basic data handling.

Mean

The sum of all data values divided by the number of values. Because every data point enters the sum, a single extreme value can shift the mean significantly.

Median

The middle value when data are arranged in order (or the average of the two middle values for even-sized sets). The median depends only on position, so extreme values rarely change it much.

Outlier

A data value that is significantly higher or lower than the rest of the data. There is no single rule, but a common test uses the interquartile range (IQR): any value more than 1.5 × IQR below Q₁ or above Q₃ is flagged.

Resistant Measure

A statistic that does not change much when outliers are added or removed. The median is resistant; the mean is not.

Skewness

When outliers pull data to one side, the distribution becomes skewed. In a right-skewed set, the mean is greater than the median; in a left-skewed set, the mean is less.

✦

KEY TAKEAWAY

Think of the mean as a balance point on a number-line seesaw. If you place a very heavy weight far to one side (the outlier), the balance point shifts toward it. The median, on the other hand, is like the person standing in the exact middle of a line: no matter how far away the last person in line wanders, the middle person barely moves.

SECTION 3

Visual Explanation — Dot Plots With and Without Outliers

The diagram below shows two dot plots using the same core data set: {2, 3, 4, 5, 6}. In the top plot, no outlier is present and the mean and median coincide at 4. In the bottom plot, the value 6 has been replaced with 20—an outlier. Notice how the mean jumps from 4 to 6.8 while the median stays at 4.

Top plot: without an outlier, mean and median both equal 4. Bottom plot: replacing 6 with the outlier value 20 pushes the mean to 6.8 while the median stays at 4. The amber arrow shows the direction and magnitude of the mean's shift.

The visual makes a critical idea concrete: because the mean uses every value in its calculation, any extremely large or small number has a direct arithmetic impact. The median ignores how far away extreme values are—it only cares about which value sits in the middle position once the data are sorted.

SECTION 4

Mathematical Framework

Let's formalize how the mean and median are calculated so you can see exactly why one is sensitive to outliers and the other is not.

ARITHMETIC MEAN

x̄ = (x₁ + x₂ + … + xₙ) / n

Where x̄ is the mean, x₁, x₂, …, xₙ are the individual data values, and n is the total number of values. Every value enters the numerator, so one huge value inflates the sum—and therefore the mean.

MEDIAN (ODD n)

Median = x₍₍ₙ₊₁₎/₂₎ (value at position (n + 1) / 2)

When n is odd, the median is the single middle value after sorting. The actual magnitudes of the smallest and largest values do not matter—only their positions do.

MEDIAN (EVEN n)

Median = (x₍ₙ/₂₎ + x₍ₙ/₂₊₁₎) / 2

When n is even, the median is the average of the two middle values. Again, outliers at the extremes have no influence unless they change which values occupy the center positions.

IQR OUTLIER TEST

Outlier if x < Q₁ − 1.5 × IQR or x > Q₃ + 1.5 × IQR

Where Q₁ is the first quartile, Q₃ is the third quartile, and IQR = Q₃ − Q₁. This is the most common method taught in introductory statistics for formally identifying an outlier.

The key insight is algebraic: in the mean formula, changing any single value directly changes the numerator by the same amount, which then propagates through the division. In the median formula, changing an extreme value has zero effect on the sorted position of the middle value. This is why statisticians call the median a resistant measure of center.

SECTION 5

Outliers, Skewness, and Choosing the Right Measure

Outliers don't just move the mean—they can change the overall shape of a distribution. When a data set has one or more extreme high values, we say the distribution is skewed right (the tail stretches to the right). With extreme low values, it is skewed left. When data are roughly symmetric, the mean and median are close together and either works well. The diagram below contrasts three distribution shapes.

In a left-skewed distribution, low outliers drag the mean below the median. In a right-skewed distribution, high outliers pull the mean above the median. In a symmetric distribution, both measures are approximately equal.

Distribution Shape	Mean vs. Median	Better Measure of Center
Symmetric (no outliers)	Mean ≈ Median	Either (mean is common)
Right-skewed (high outliers)	Mean > Median	Median
Left-skewed (low outliers)	Mean < Median	Median

Guidelines for selecting mean or median based on distribution shape.

SECTION 6

Worked Example — Test Scores

A class of nine students takes a quiz. The scores are: 72, 75, 78, 80, 82, 84, 85, 88, 150. The score of 150 looks suspiciously high (perhaps the teacher entered the wrong grade). Let's compute the mean and median, then decide which one better represents the typical student score.

Quiz Score Analysis

1
Step 1 — Sort the Data

The data are already sorted from least to greatest: 72, 75, 78, 80, 82, 84, 85, 88, 150.
2
Step 2 — Find the Median

There are n = 9 values (odd), so the median is the value in position (9 + 1) / 2 = 5th position. Counting: 72, 75, 78, 80, 82, 84, 85, 88, 150.

Median = 82
3
Step 3 — Calculate the Mean

Sum = 72 + 75 + 78 + 80 + 82 + 84 + 85 + 88 + 150 = 794. Mean = 794 / 9 ≈ 88.2.

Mean ≈ 88.2
4
Step 4 — Identify the Outlier

The value 150 is far above the other scores. We can verify using the IQR method. Q₁ = 76.5, Q₃ = 86.5, so IQR = 10. Upper fence = 86.5 + 1.5 × 10 = 101.5. Since 150 > 101.5, the value 150 is a confirmed outlier.
5
Step 5 — Compare and Justify

The mean of 88.2 is higher than 8 out of 9 actual scores—it does not represent a typical student. The median of 82 sits right in the middle of the non-outlier scores. Because the data contain an outlier that skews the distribution to the right, the median is the better measure of center for this data set.

Best measure: Median = 82

SECTION 7

Strengths & Limitations — Mean vs. Median

Feature	Mean	Median
Sensitivity to outliers	High — one extreme value changes it significantly	Low — remains stable even with extreme values (resistant)
Uses all data values	Yes — every value contributes to the sum	No — only the middle value(s) matter
Best for symmetric data	Excellent — equals the median and is easy to compute	Also good, but provides the same answer as the mean
Best for skewed data	Misleading — pulled toward the tail	Preferred — stays near the bulk of the data
Common real-world use	GPA, batting average, temperature averages	Household income, home prices, hospital wait times

Comparison of the mean and median as measures of center.

💡

REAL-WORLD EXAMPLE

News reports about household income in the United States almost always use the median rather than the mean. Why? A small number of billionaires earn so much that the mean household income is pulled far above what a typical family actually earns. The median gives a more honest picture of the 'middle' American household. When you see 'average' in a headline, always ask: is that the mean or the median?

SECTION 8

Connection to Advanced Statistics

The ideas you've learned here extend into more advanced courses in surprising ways. Understanding how outliers affect the center prepares you for topics like robust regression, trimmed means, and non-parametric statistics—all of which deal with data that don't follow neat, symmetric patterns.

Concept in This Lesson	Advanced Extension
Mean is sensitive to outliers	Trimmed mean (drop top/bottom 5–10% of data) combines resistance with the mean's mathematical usefulness
Median is resistant	Median Absolute Deviation (MAD) extends resistance to measures of spread
IQR rule for outliers	Z-score method (\|z\| > 2 or 3) is used in AP Statistics and beyond for more formal outlier detection
Choosing mean vs. median	In inference, non-parametric tests like the Wilcoxon rank-sum test use medians instead of means when assumptions are violated

How today's concepts connect to future coursework.

For now, the most important skill is being able to justify your choice of measure. On tests and projects, stating 'the median is better because outliers pull the mean away from the center of the data' earns full credit. In more advanced courses, you'll learn formal ways to quantify how much influence an outlier has.

SECTION 9

Practice Problems

Question 1: CONCEPTUAL

Explain in your own words why the median is called a "resistant" measure of center. What does it resist, and why?

Show answer

The median is called resistant because it does not change much—or at all—when extreme values (outliers) are added to or removed from a data set. It resists the pull of outliers because it depends only on the middle position of the sorted data, not on the actual magnitudes of the extreme values. For example, whether the largest value in a five-element set is 10 or 10,000, the median remains the third value in sorted order.

Question 2: BASIC CALCULATION

Find the mean and median of the data set: {10, 12, 14, 15, 16, 18, 50}. Identify the outlier and state which measure better represents the typical value.

Show answer

Sum = 10 + 12 + 14 + 15 + 16 + 18 + 50 = 135. Mean = 135 / 7 ≈ 19.3. The median is the 4th value (since n = 7): Median = 15. The value 50 is an outlier (it is far from the cluster of 10–18). The mean of 19.3 is higher than 6 out of 7 values, so it overstates the typical value. The median of 15 better represents the center of this data set because it is not pulled by the outlier.

Question 3: INTERMEDIATE

A data set has a mean of 40 and a median of 40. You add a new value of 200 to the set. Without calculating exact values, describe the effect on the mean and the median. Which measure changes more? Justify your answer.

Show answer

Adding the value 200 will increase the mean noticeably because the sum increases by 200 while n increases by only 1, so the new mean will be pulled upward toward 200. The median will change very little because the new value goes to the far right of the sorted list; the middle position shifts by at most half a position, which typically results in only a small change (or none). Therefore the mean changes more. This is because the mean uses every data point in its calculation, making it non-resistant to outliers, while the median relies only on positional order.

Question 4: APPLIED

The following values represent the sale prices (in thousands of dollars) of eight homes in a neighborhood: 220, 235, 240, 250, 255, 260, 270, 980. A real estate agent wants to advertise the 'average' home price. Calculate the mean and median. Which measure should the agent use to give buyers a realistic expectation, and why?

Show answer

Sum = 220 + 235 + 240 + 250 + 255 + 260 + 270 + 980 = 2,710. Mean = 2,710 / 8 = 338.75 (i.e., $338,750). Median = (250 + 255) / 2 = 252.5 (i.e., $252,500). The value $980,000 is an outlier pulling the mean far above most sale prices. Only one home sold above $338,750, so the mean gives an inflated expectation. The median of $252,500 sits right in the middle of the seven non-outlier prices and gives buyers a much more realistic picture of what a typical home costs. The agent should use the median.

Question 5: CRITICAL THINKING

Samira claims, 'If a data set has no outliers, the mean and median will always be exactly equal.' Is her claim correct? Provide a counterexample if it is false, and explain the relationship between outliers and the equality of mean and median.

Show answer

Samira's claim is incorrect. A data set can have no outliers yet still be slightly skewed, causing the mean and median to differ. For example, consider {1, 2, 3, 4, 10}. Using the IQR test: Q₁ = 1.5, Q₃ = 7, IQR = 5.5, upper fence = 7 + 1.5 × 5.5 = 15.25. Since 10 < 15.25, there is no outlier. However, the mean = 20 / 5 = 4.0 while the median = 3. They are not equal because the data are mildly right-skewed. The key idea is that outliers guarantee the mean will be pulled away from the median, but the absence of outliers does not guarantee the two will be equal—it only makes them closer together.

SUMMARY

Lesson Summary

Summary

An outlier is a data value that is far removed from the rest of the data. Because the mean uses every value in its calculation (sum ÷ count), a single extreme number can pull the mean significantly toward itself. The median, defined as the middle value of sorted data, depends only on position and is therefore a resistant measure of center that remains stable despite outliers.

When a distribution is skewed by outliers, the median is usually the better measure of center because it more accurately represents the typical value. In symmetric distributions without outliers, the mean and median are approximately equal and either may be used. To justify your choice on an assessment, state whether outliers are present, describe how they affect the mean, and explain why the median is or is not a better representation of the data's center.