Grouped frequency table and relative frequency chart

MathematicsStatisticsAges 14–15

Use with my class

A data sample (student height, weight, or daily self-study time) is grouped into classes of a chosen width. The frequency table n_i, relative frequency f_i = n_i/n·100%, and the histogram update instantly as you drag the width slider; click a bar to see the values in that class, or click "New data" for a fresh sample. Two more screens: Frequency density (unequal class widths you can drag; bar height = frequency ÷ class width) and Cumulative frequency (read the median, quartiles, interquartile range and any percentile from the graph, with a box plot).

Lesson: Grouped frequency tables, grouped relative frequency, and histograms

What it shows

With equal classes [a; a + w) the simulation counts the frequency n_i and the relative frequency f_i = n_i/n · 100%. When classes have unequal widths, the bar height must be the frequency density = frequency ÷ class width, so that the area of each bar equals its frequency. The cumulative frequency graph shows how many values lie below each upper class boundary; reading across at n/4, n/2 and 3n/4 gives estimates of Q1, the median and Q3, which build a box plot. For box plots of raw data, see the sim "Summary statistics of a data set".

How to use

On Equal classes, drag the Class width slider and switch between Frequency and Relative frequency. On Frequency density, drag the ▲ handles to move the class boundaries, click a bar to see f ÷ w, and press Height = frequency (misleading) to compare. On Cumulative frequency, move the Percentile P slider, tap the graph to read how many values lie below x, and compare the graph estimates with the raw-data median and quartiles.

Parameters you can change

  • Sample size n 20–200
  • Class width 2–20
  • Chart quantity Frequency n_i, Relative frequency f_i (%)
  • Data set Height of Grade 9 students (cm), Weight of students (kg), Daily self-study time (minutes)
  • Show frequency polygon
  • Screen Equal classes, Frequency density (unequal classes), Cumulative frequency and box plot
  • Percentile P (Cumulative frequency screen) 1–99

Questions to explore

  1. What do the relative frequencies in the table always add up to, whatever class width you choose?
  2. Which class width gives the clearest picture of the distribution for this data set, and why?
  3. Into which class does a value lying exactly on the boundary between two classes go?