Histograms: Seeing the Shape of Data
Imagine you have the marks of 50 students in a class. The raw list — 67, 43, 81, 55, 72, 39, 91, 48, 63, 77, and so on — is just a jumble of numbers. You can't easily tell: Are most students scoring around 70? Is the class weak in a particular range? How spread out are the marks?
A histogram is a picture that answers these questions at a glance. It takes the data, groups it into intervals (called bins or class intervals), and draws a bar for each interval. The height of the bar tells you how many data points fall into that interval.
A histogram looks like a bar chart, but it is not a bar chart. In a bar chart, the bars are for categories (like "Apple", "Banana", "Orange") — you can rearrange them. In a histogram, the bars are for number ranges (like "0–10", "10–20") — the order is fixed by the numbers themselves.
The Intuition: Stacking Blocks
Think of a number line from 0 to 100. Take each student's mark and place a block above that exact number. You'll get a messy scatter of blocks. Now, instead of placing each block at its exact point, you decide to group them: "All marks from 0 to 10 go in the first pile, 10 to 20 in the second pile, …" You stack the blocks for each pile. The height of each stack is the frequency — the number of students in that range.
That stack of blocks is your histogram. The horizontal axis shows the intervals (the bins), and the vertical axis shows the frequency (the count).
The Precise Definition
A histogram is a graphical representation of a frequency distribution for continuous (or grouped) data. It consists of:
- Class intervals (bins) on the horizontal axis. These are consecutive, non-overlapping intervals of equal width (usually).
- Rectangles (bars) drawn on each interval. The area of each rectangle is proportional to the frequency of that interval. When the intervals are of equal width, the height of the rectangle is directly proportional to the frequency.
Area of bar∝Frequency of that class
For equal-width classes: Height of bar∝Frequency
How to Draw a Histogram (Step by Step)
Let's use the marks of 50 students, grouped into intervals of width 10:
| Marks (Class Interval) | Frequency (Number of Students) |
|---|
| 0 – 10 | 2 |
| 10 – 20 | 3 |
| 20 – 30 | 5 |
| 30 – 40 | 8 |
| 40 – 50 | 12 |
| 50 – 60 | 10 |
| 60 – 70 | 6 |
| 70 – 80 | 3 |
| 80 – 90 | 1 |
| 90 – 100 | 0 |
Step 1: Draw a horizontal axis. Mark the class intervals continuously: 0, 10, 20, 30, …, 100. There are no gaps between intervals — the bar for 0–10 touches the bar for 10–20.
Step 2: Draw a vertical axis. Label it "Frequency". Choose a scale that fits your highest frequency (here, 12).
Step 3: For each interval, draw a rectangle whose base spans the interval and whose height reaches the frequency. For the interval 40–50, the rectangle goes from 40 to 50 on the horizontal axis and up to 12 on the vertical axis.
Step 4: Label the axes and give the graph a title (e.g., "Distribution of Marks in a Class Test").
In exams, always check whether the class intervals are continuous (like 0–10, 10–20) or discontinuous (like 0–10, 11–20). If they are discontinuous, you must first make them continuous by adjusting the boundaries. For example, 0–10 and 11–20 become 0–10.5 and 10.5–20.5 (or 0–10, 10–20 by subtracting 0.5 from the lower limit and adding 0.5 to the upper limit).
What a Histogram Tells You
- Shape of the data: Is it symmetric (bell-shaped), skewed to the left (more high values), or skewed to the right (more low values)?
- Central tendency: Which interval has the highest bar? That's the modal class — the range where most data lies.
- Spread: How wide is the distribution? Are the bars concentrated in a few intervals or spread across many?
- Outliers: Are there isolated bars far from the main cluster?
The Critical Difference: Histogram vs. Bar Chart
| Feature | Histogram | Bar Chart |
|---|
| Data type | Continuous (numbers) | Categorical (labels) |
| Bar order | Fixed by number order | Can be rearranged |
| Bar gaps | No gaps between bars | Gaps between bars |
| Bar width | Represents the interval width | Arbitrary (same width) |
| Area | Proportional to frequency | Height = frequency |
A Common Mistake
Students often draw gaps between the bars of a histogram. Do not. The data is continuous — there is no break between 10 and 20. The bars must touch. If you leave gaps, you are implying that the data is categorical, which defeats the purpose of a histogram.
Never leave gaps between bars in a histogram. The only exception is when a class interval has zero frequency — then you still show the bar (of height zero) as a line on the axis, but you do not skip the interval.
Why It Matters
Histograms are the first tool statisticians use to "see" data. Before calculating averages or standard deviations, a histogram shows you the big picture. In your exams, you will be asked to draw histograms from frequency tables, or to interpret a given histogram to find the modal class, estimate the mean, or compare two distributions. Master the histogram, and you master the first step of data analysis.