Histograms — Frequency, Class Width & How to Read

A histogram shows how often numerical values fall into intervals such as $0$ to $10$ or $10$ to $20$. The class width is the size of each interval, and the frequency is how many values land in it. Reading one well is a short, repeatable procedure rather than a single glance.

When To Use This Reading Method

Reach for histogram reading when you have numerical data grouped into ranges, often called classes or bins, and you want the shape of the data rather than individual values. The bars touch because the intervals sit next to each other on a number line.

That is also why a histogram is not a bar chart. A bar chart compares separate categories such as sports or colors; a histogram shows the shape of a distribution. If your data is categorical, this method does not apply.

The Step-By-Step Procedure

Step 1 — Read the horizontal axis. Start with the intervals so you know what each bar actually covers, and note the class width of each one.

Step 2 — Check whether the class widths are equal. This decides everything that follows.

Step 3 — If widths are equal, compare heights directly. The tallest bars mark the most common intervals. The frequency of a class is the number of observations in that interval; if the class $60$ to $70$ contains $8$ test scores, its frequency is $8$.

Step 4 — If widths are not equal, switch to area. Do not compare heights automatically. In many courses the vertical axis becomes frequency density, so bar area represents frequency:

Step 5 — Scan the overall shape. Find the center, spot the gaps, and notice whether one side stretches farther than the other.

The Whole Procedure On One Example

Suppose a histogram summarizes these quiz scores, each class with width $10$:

Reading the horizontal axis (Step 1) shows five intervals of width $10$, so the widths are equal (Step 2) and bar heights can be compared directly (Step 3). The tallest bar is $60$ to $70$, so that interval contains the most scores. Scanning the shape (Step 5): most scores fall between $50$ and $80$, and only a few are below $50$ or above $80$.

A clean summary: the scores cluster in the middle, with the biggest concentration between $60$ and $70$.

Where Each Step Trips People Up, And How To Self-Check

• At Step 2 — mistaking a histogram for a bar chart. In a histogram the bars touch because the intervals connect; in a bar chart the categories are separate, so gaps are normal. Self-check: do the bars touch?
• At Step 3 — ignoring class width. Comparing heights only works when the widths are equal or the axis already uses frequency density. Self-check: are all the widths the same number?
• At Step 1 — handling endpoints carelessly. Grouped data needs a consistent rule about class boundaries; a score of $70$ should belong to one class, not both. The labeling or context tells you which side is included.
• After Step 5 — expecting exact raw data. A histogram summarizes grouped data. It shows the pattern well but does not let you recover every original value the way a stem-and-leaf plot can.

When This Is Worth Doing

Histograms are useful when you want a quick picture of how numerical data is distributed: test scores, response times, science-lab readings, quality-control data. They are especially helpful before computing summary statistics, because they show whether the data looks balanced, skewed, clustered, or unusually spread out.

To practice the full procedure, take a small set of numerical data, sort it into equal-width intervals, sketch the histogram by hand, and write one sentence describing the pattern before you compute the mean or median. Repeat once with unequal widths and watch how the picture, and Step 4, change.