Descriptive StatisticsTopic #1 of 33

Mean, Median, and Mode

Central tendency measures: arithmetic mean, median for ordered data, and mode for most frequent values.

Overview

The mean, median, and mode are three measures of central tendency—ways to describe the "center" of a dataset.

Definitions

MeasureDefinitionWhen to Use
MeanArithmetic averageSymmetric data, no outliers
MedianMiddle valueSkewed data, ordinal data
ModeMost frequent valueCategorical data, bimodal

Formulas

Arithmetic Mean

For a dataset with nn values:

xˉ=x1+x2++xnn=xin\bar{x} = \frac{x_1 + x_2 + \cdots + x_n}{n} = \frac{\sum x_i}{n}

Population Mean

μ=xiN\mu = \frac{\sum x_i}{N}

Weighted Mean

xˉw=(wi×xi)wi\bar{x}_w = \frac{\sum (w_i \times x_i)}{\sum w_i}

Median Calculation

  1. Order the data from smallest to largest
  2. If nn is odd: median = middle value
  3. If nn is even: median = average of two middle values
Median position=n+12\text{Median position} = \frac{n + 1}{2}

Examples

Odd number of values (5 values): 2, 5, 7, 9, 12

  • Position: (5+1)/2=3(5+1)/2 = 3
  • Median = 7 (the 3rd value)

Even number of values (6 values): 2, 5, 7, 9, 12, 15

  • Positions: 3rd and 4th values
  • Median = (7+9)/2=8(7 + 9)/2 = 8

Mode

The value(s) that appear most frequently.

TypeDescription
UnimodalOne mode
BimodalTwo modes
MultimodalMultiple modes
No modeAll values appear equally

Properties

Mean Properties

  • Uses all data values
  • Affected by outliers
  • Sum of deviations from mean = 0: (xixˉ)=0\sum (x_i - \bar{x}) = 0
  • Can be calculated for interval/ratio data

Median Properties

  • Not affected by outliers
  • 50% of data above, 50% below
  • Can be calculated for ordinal, interval, or ratio data

Mode Properties

  • Can be used for categorical data
  • Not affected by outliers
  • May not exist or may not be unique

Relationship in Distributions

DistributionRelationship
SymmetricMean ≈ Median ≈ Mode
Right-skewedMode < Median < Mean
Left-skewedMean < Median < Mode

Choosing the Right Measure

SituationBest Measure
Normal distributionMean
Skewed distributionMedian
Outliers presentMedian
Categorical dataMode
Ordinal dataMedian or Mode

Examples

Example 1: Test Scores

Data: 85, 90, 78, 92, 88, 85, 95, 85

Mean=85+90+78+92+88+85+95+858=6988=87.25\text{Mean} = \frac{85 + 90 + 78 + 92 + 88 + 85 + 95 + 85}{8} = \frac{698}{8} = 87.25

Ordered: 78, 85, 85, 85, 88, 90, 92, 95

Median=85+882=86.5\text{Median} = \frac{85 + 88}{2} = 86.5 Mode=85 (appears 3 times)\text{Mode} = 85 \text{ (appears 3 times)}

Example 2: With Outlier

Salaries: $40k, $42k, $45k, $48k, $200k

Mean=40+42+45+48+2005=$75k\text{Mean} = \frac{40 + 42 + 45 + 48 + 200}{5} = \$75\text{k} Median=$45k (3rd value)\text{Median} = \$45\text{k} \text{ (3rd value)}

The median ($45k) better represents typical salary.

Example 3: Weighted Mean

Grades: Homework (20%): 90, Midterm (30%): 80, Final (50%): 85

xˉw=(0.20×90)+(0.30×80)+(0.50×85)0.20+0.30+0.50\bar{x}_w = \frac{(0.20 \times 90) + (0.30 \times 80) + (0.50 \times 85)}{0.20 + 0.30 + 0.50} =18+24+42.51=84.5= \frac{18 + 24 + 42.5}{1} = 84.5