Graphing frequency starts with organizing your raw data into a table that counts how often each value or range of values appears, then choosing the right chart type to display that count visually. The x-axis represents your data categories or ranges, and the y-axis represents how many times each one occurs. The process changes slightly depending on whether your data is continuous (like heights or temperatures) or categorical (like favorite colors or survey responses), so picking the right graph type is your first real decision.
Organize Your Data First
Before you touch a graph, you need a frequency distribution table. This is simply a list of every possible value or range paired with a count of how many times it shows up in your data set. For small, categorical data (like types of pets owned by 56 people), you list each category in one column and the count in the next. A tally column in the middle helps you track counts as you work through raw data by hand.
For continuous data like test scores or ages, you’ll need to group values into class intervals (also called bins). If your data ranges from 0 to 100, you might create intervals of 0–9, 10–19, 20–29, and so on. The number of intervals matters: too few and you lose detail, too many and the graph becomes noisy. A common starting point is to use somewhere between 5 and 20 bins depending on your data set size. Once your intervals are set, count how many data points fall into each one. That count is your frequency.
Choose the Right Graph Type
Four main graph types handle frequency data, and each one fits a different situation.
- Histogram: Best for continuous data grouped into intervals. The x-axis shows the ranges, the y-axis shows frequency, and the bars touch each other to signal that the data flows across a continuous scale. Bar widths can vary if your class intervals aren’t equal, because in a histogram the area of each bar (not just its height) represents the frequency.
- Bar chart: Best for categorical or discrete data, like the number of votes each candidate received. The bars are separated by gaps to emphasize that each bar represents a distinct, independent category. All bars are equal width, and height alone represents the count.
- Frequency polygon: A line graph created by plotting a point at the midpoint of each class interval at the height of its frequency, then connecting those points with straight lines. This is especially useful when you want to compare two distributions on the same graph, since overlapping lines are easier to read than overlapping bars.
- Ogive (cumulative frequency graph): Shows a running total of frequencies. Instead of plotting how many values fall in each interval, you plot how many values fall at or below the upper boundary of each interval. The result is a curve that always rises, letting you quickly read off how much of your data falls below any given value.
How to Build a Histogram Step by Step
Start by determining your class intervals. Write them along the x-axis from smallest to largest with no gaps between them. Then draw the y-axis tall enough to accommodate your highest frequency count. For each interval, draw a bar whose height matches the number of data points in that range. The bars should be flush against each other, with no space in between.
Label both axes clearly. The x-axis should name the variable you’re measuring (for example, “Test Score”) and the y-axis should say “Frequency” or “Count.” Add a descriptive title above or below the graph. Use a readable sans serif font if you’re preparing this for a report or presentation, with text sized between 8 and 14 points. Including units of measurement on the axes is a standard expectation in any professional or academic setting.
How to Build a Frequency Polygon
A frequency polygon uses the same underlying data as a histogram but represents it with connected points instead of bars. First, calculate the midpoint of each class interval by adding the lower limit to the upper limit and dividing by 2. For an interval of 20–29, the midpoint is 24.5.
Plot each midpoint on the x-axis and its corresponding frequency on the y-axis. These midpoints are your x-coordinates and the frequencies are your y-coordinates. Then connect each point to the next with a straight line segment, working left to right. You can optionally extend the line down to the x-axis at both ends (using the midpoints of the hypothetical intervals before your first and after your last) to close the polygon.
This format shines when you’re comparing two data sets. Plot both polygons on the same axes using different colors or line styles, and the overlaps and differences become immediately visible.
How to Build an Ogive
An ogive graphs cumulative frequency, meaning you’re plotting a running total rather than individual counts. To build one using the “less than” method, start by calculating the cumulative frequency for each class interval. For the first interval, the cumulative frequency is just its own count. For the second, add the first and second counts together. Continue adding each new interval’s count to the running total.
Plot each cumulative frequency against the upper boundary of its class interval. If your intervals are 0–9, 10–19, and 20–29 with frequencies of 5, 12, and 8, you’d plot (9, 5), (19, 17), and (29, 25). Connect the points with straight lines or a smooth freehand curve. The result always slopes upward, and you can use it to estimate percentiles or medians by reading across from the y-axis to the curve and then down to the x-axis.
There’s also a “more than” method, where you subtract each interval’s frequency from the total and plot against the lower class boundary. This produces a downward-sloping curve. Where the two ogives cross, if plotted together, gives you the median of the distribution.
Graphing Relative Frequency
Sometimes raw counts aren’t as useful as proportions or percentages. Relative frequency tells you what fraction of your total data set falls in each category or interval. To calculate it, divide the frequency of each group by the total number of data points. If 16 out of 56 people own dogs, the relative frequency is 16 ÷ 56 = 0.29, or 29%.
You can express relative frequency as a decimal, a fraction, or a percentage. Any of these is technically correct, though decimals and percentages are the most common choices. To graph relative frequency, use the same chart types described above but replace the y-axis label with “Relative Frequency” or “Percentage” instead of “Count.” The shape of the graph stays identical to the raw frequency version; only the scale changes. This makes it possible to compare distributions from data sets of different sizes, since everything is normalized to proportions.
Common Mistakes to Avoid
The most frequent error is using a histogram for categorical data or a bar chart for continuous data. If your x-axis categories have a natural order and no gaps between them (age ranges, income brackets, time intervals), use a histogram with touching bars. If the categories are independent and could be rearranged in any order (countries, product names, survey options), use a bar chart with spaced bars.
Another common problem is choosing too many or too few bins for a histogram. With too few bins, you mask the real shape of your distribution. With too many, every bar is short and the graph looks like static. Start with a moderate number and adjust until the overall pattern is clear without excessive noise.
Unlabeled axes are a third issue. Every frequency graph needs a clearly labeled x-axis identifying the variable, a y-axis labeled “Frequency” (or “Relative Frequency” or “Cumulative Frequency” depending on the type), and a title that tells the reader exactly what they’re looking at. Skipping any of these forces the reader to guess, which defeats the purpose of visualizing the data in the first place.