Cumulative frequency and histograms

For large sets of grouped data, cumulative frequency (a running total) lets you estimate the median and quartiles from a curve. Histograms show grouped data with unequal class widths, where area shows frequency.

Key facts

  • Cumulative frequency: add up the frequencies as you go.
  • Plot at the upper end of each class, then join with a smooth curve.
  • Median at n/2, Q1 at n/4, Q3 at 3n/4 (read across, then down).
  • Histogram: frequency density = frequency ÷ class width; area = frequency.
  • Estimated mean from grouped data uses midpoints.

Worked examples

Example 1

A class of 0–10 has frequency 12 and 10–30 has frequency 30. Find the frequency densities.

  1. 0–10: 12 ÷ 10 = 1.2
  2. 10–30: 30 ÷ 20 = 1.5

Answer: 1.2 and 1.5

Example 2

Estimate the mean: 0–10 (f = 4), 10–20 (f = 10), 20–30 (f = 6).

  1. Midpoints 5, 15, 25.
  2. Σfx = 20 + 150 + 150 = 320, Σf = 20
  3. 320 ÷ 20 = 16

Answer: 16