VibeMath

Statistics and linear algebra

Mean, median, and the effect of outliers

Mean and median both describe a center, but they respond differently to extreme values. The mean redistributes every value into an equal share, while the median depends mainly on order and the middle position.

Author
VibeMath Editorial TeamOriginal AI-assisted lesson content edited for clarity, accessibility, and product-truth accuracy.
Math reviewer
VibeMath Math ReviewIndependent equation, example, substitution, units, scope, and accessibility checks; not a claim of individual human credentials.
Last reviewed

What you will learn

  • Interpret the mean as an equal-share or balance-point value.
  • Find the median from ordered data with odd or even sample size.
  • Explain why a large outlier usually moves the mean more than the median.

Build the idea

The mean uses every observation. Add the values and divide by how many there are. One visual interpretation treats each value as a stack of blocks: if blocks could be moved between stacks until all heights matched, the common height would be the mean. A distant value contributes many blocks and can pull that balance point toward itself.

The median begins by ordering the observations. With an odd count, it is the single middle value. With an even count, it is halfway between the two middle values. Changing an extreme observation without moving it past the center may leave those middle positions unchanged, so the median is resistant to the size of an outlier.

Neither measure is automatically better. The mean is informative when the distribution is reasonably balanced and every numerical distance matters. The median is often more representative for a strongly skewed distribution, but it deliberately ignores how far the extremes lie from the center. A responsible summary names the choice and inspects the full distribution.

Stacks of data blocks and a movable center marker

Show five similarly sized stacks at four, five, six, seven, and eight, then redistribute blocks so each stack reaches six. Add a sixth stack at thirty and show that equal sharing rises to ten while the two middle markers remain at six and seven.

Visual description: Ordered dots at four through eight have mean and median six. Adding a distant dot at thirty moves the mean to ten, while the median becomes six point five near the original cluster.
Read as: x bar equals x one plus x two and the remaining values, all divided by n.
Read as: For an even count, the median equals the two middle ordered values added and divided by two.

Worked example

Problem

Compare mean and median for four, five, six, seven, and eight, then add the outlier thirty and describe how each center changes.

Strategy

Calculate both summaries before and after the additional observation, keeping the values ordered so positional and arithmetic effects remain visible.

  1. Step 1

    The five original values sum to thirty. Dividing by five gives an equal-share value of six, located at the center of this symmetric set.

    Read as: Four plus five plus six plus seven plus eight, divided by five, equals six.

  2. Step 2

    There are five ordered values, so the third value is the unique middle observation. It is also six, matching the mean before the outlier appears.

    Read as: The median of four, five, six, seven, and eight equals six.

  3. Step 3

    After adding thirty, the total becomes sixty across six observations. Equal sharing moves the mean upward to ten, even though five of the six values remain at eight or below.

    Read as: Four plus five plus six plus seven plus eight plus thirty, divided by six, equals ten.

  4. Step 4

    With six ordered values, average the third and fourth positions. Those values are six and seven, so the median shifts only one half unit to six point five.

    Read as: The quantity six plus seven, divided by two, equals six point five.

Answer and verification

The outlier changes the mean from six to ten, while the median changes from six to six point five.

The new mean must preserve the total: six groups of ten account for all sixty units. The new median must split the ordered list into three observations on each side of six point five. Both checks succeed and illustrate the different definitions.

Common mistakes

  • Choosing the mean because it uses more arithmetic, without examining whether an extreme value dominates the result.

    Plot or sort the distribution before choosing a center. If a distant value pulls the mean away from most observations, report the median or report both measures with an explanation.

  • Taking one of the two middle values as the median when the data set has an even number of observations.

    For an even count, there is no single middle observation. Average the two central ordered values so the median lies halfway between them.

Try it yourself

Find the mean and median of two, three, four, five, and sixteen, then explain which measure better reflects the cluster from two through five.

Show hint

Order is already given; calculate the equal-share value and compare it with the single middle position.

Show answer

The mean is six and the median is four, so the median lies closer to the four-value cluster when sixteen is treated as an outlier.

Accessible lesson transcript

This reviewed reading order covers the lesson explanation, equations, example, and checks without claiming that a video exists.

  1. Part 1

    Compare two meanings of center

    The mean and median answer different questions. The mean asks what equal share would preserve the total, so it uses every value and every distance. The median asks where the ordered data split in half, so it depends on positions near the center. We will watch those definitions respond to one distant observation.

  2. Part 2

    Start with a balanced set

    For four, five, six, seven, and eight, the total is thirty and the mean is six. Six is also the third of five ordered values, so it is the median. The two summaries coincide because the data are evenly arranged around six, not because mean and median are interchangeable.

  3. Part 3

    Introduce one distant value

    Add thirty as a sixth observation. The total doubles to sixty even though the observation count rises only from five to six. Equal sharing therefore moves the mean to ten. The value thirty contributes twenty units above the new mean, exactly balancing the other five observations' total deficit below ten.

  4. Part 4

    Track the middle positions

    The ordered middle values are still six and seven, so the new median is six point five. It moves only slightly while the mean moves four units. This does not make median universally correct. It shows why median often describes a skewed cluster more faithfully, while mean still carries information about the total and extreme magnitude.

Scope and limitations

  • This lesson uses small unweighted data sets and does not cover grouped data, sampling uncertainty, weighted means, or formal outlier-detection rules.
  • Labeling a value an outlier requires subject-matter context or a stated statistical rule; distance alone does not prove that an observation is erroneous.