Basic Statistics Without Fear: Mean, Median, Mode

Three Words That Change Everything: Mean, Median, Mode

"The average American household earns $105,000 per year." That statement, if you read it recently, might have made you feel either reassured or suspicious. You should feel suspicious — because this uses the mean, which is heavily distorted by the ultra-wealthy, while the median (the typical middle household) earns closer to $74,000. The choice of which "average" to report is a powerful tool that shapes how we understand inequality, health outcomes, real estate prices, and almost every domain of data.

Percentage

Mean: The Arithmetic Average

The mean is calculated by summing all values and dividing by the count:

Mean = (x₁ + x₂ + x₃ + ... + xₙ) / n

Example: Five salaries in a small company: $45,000, $50,000, $52,000, $55,000, $550,000 - Sum = $752,000 - Mean = $752,000 / 5 = $150,400

This mean is heavily influenced by the CEO's $550,000 salary. The "average" salary of $150,400 does not represent any actual employee's experience — four out of five employees earn between $45k and $55k.

When mean is appropriate: - When the distribution is roughly symmetric (no extreme outliers) - When calculating totals matters (e.g., total revenue = mean revenue × customers) - When performing further statistical calculations (mean is mathematically tractable in ways median is not)

Median: The Middle Value

The median is the value that falls exactly in the middle when data is sorted. If there is an even number of values, it is the mean of the two middle values.

Sorted salaries: $45,000 | $50,000 | $52,000 | $55,000 | $550,000 - Median = $52,000

The median is the salary a randomly selected employee is most likely to earn (within a reasonable range). It is immune to the distorting effect of outliers.

When median is appropriate: - Income and wealth distributions (highly skewed by the ultra-wealthy) - House prices (skewed by luxury properties) - Test scores with floor or ceiling effects - Any dataset with extreme outliers that don't represent the typical case

Median home price vs. mean home price: In a neighborhood with 99 houses worth $300,000 and one mansion worth $20 million, the mean is $497,000 but the median is $300,000. Which better represents what a typical buyer would pay? The median, clearly.

Weighted Average Formula

Mode: The Most Common Value

The mode is the value that appears most frequently. A dataset can have: - No mode (all values appear equally often) - One mode (unimodal) - Two modes (bimodal) - Multiple modes (multimodal)

Example: Shoe sizes sold in one day: 7, 8, 8, 9, 9, 9, 10, 10, 11, 12 - Mode = 9 (appears 3 times)

When mode is appropriate: - Categorical data where mean/median don't apply (most popular color, most common answer choice) - Inventory decisions (stock the mode size in greatest quantity) - Identifying the most common failure mode in quality control

Mode is rarely used in isolation for continuous numerical data but matters enormously for categorical analysis.

The Shape of Data Distributions

Understanding why mean and median diverge requires understanding distribution shape:

Symmetric distribution (e.g., heights, IQ scores): - Mean = Median = Mode - The classic bell curve (normal distribution) - In this case, all three measures of center agree

Right-skewed (positively skewed) distribution (e.g., incomes, wealth): - A long tail extends to the right (high values) - Mean > Median (mean is pulled right by extreme values) - Most individuals are below the mean - This is why income statistics should typically report median

Left-skewed (negatively skewed) distribution (e.g., exam scores when most students do well): - A long tail extends to the left (very low values) - Mean < Median - Most individuals are above the mean

Real Estate Prices: A Case Study

Real estate markets provide a constant demonstration of mean vs. median in action.

Consider a city block with these sale prices: $280,000, $295,000, $310,000, $320,000, $3,400,000 (the corner mansion). - Mean: ($280k + $295k + $310k + $320k + $3,400k) / 5 = $921,000 - Median: $310,000

Real estate reports use median for exactly this reason. "Median home price rose 3.2%" is a meaningful statement about the middle of the market. "Mean home price rose 3.2%" is distorted by what happened to the most expensive properties.

Standardized Test Scores: When Percentiles Beat Averages

For standardized tests like the SAT, ACT, or GMAT, the most useful measure is neither mean nor median but percentile rank — where you fall in the distribution relative to all test-takers.

A mean SAT score of 1060 (approximately the national mean) tells an applicant nothing about competitiveness at specific schools. Knowing that 1060 is the 52nd percentile, while the school's middle 50% range is 1350–1530, is immediately actionable.

Standard deviation: Measures how spread out the data is around the mean. - SAT: mean ≈ 1060, standard deviation ≈ 215 - 68% of test-takers score within one standard deviation of the mean (845–1275) - 95% score within two standard deviations (630–1490)

A score of 1490 (mean + 2 SDs) is in approximately the 98th percentile. Standard deviation gives mean context — a mean of 75 with SD = 5 indicates tight clustering, while SD = 25 indicates wide variability.

Practical Statistical Literacy

Four questions to ask whenever you see a reported "average":

1. Which average? If not specified, ask whether mean or median is more appropriate
2. What is the distribution shape? Is it skewed? Are there outliers?
3. How many data points? A mean of "5 customer reviews" is meaningless
4. What is the variability? A mean salary of $70,000 with standard deviation $5,000 describes a very different situation from mean $70,000 with standard deviation $40,000

These questions transform you from a passive consumer of statistics to an active interpreter — a genuinely valuable skill in a world drowning in data.

Percentage