Histograms and Box Plots

Histograms show the shape of a distribution — where values cluster, how spread they are, and whether they skew. Box plots summarise that distribution in five numbers and make outliers instantly visible.

Distribution — The Question Behind Both Charts

Bar charts compare totals. Line charts show trends. Histograms and box plots answer a different question entirely: across all your data points, how are the values spread out? Are most orders worth a similar amount, or do a handful of giant orders skew the total? Are delivery times consistently fast, or wildly variable? Distribution charts expose the shape behind the summary statistic.

A SUM or AVG alone can be deeply misleading. An average order value of $450 could mean every order is roughly $450 — or it could mean 90% of orders are under $100 and a few massive orders drag the average up. Only a distribution chart reveals which situation you are actually in.

Histograms — Frequency Across Value Ranges

A histogram divides a continuous Measure into equal-width buckets called bins and counts how many data points fall into each bin. The X axis shows the value range, the Y axis shows the count of records in that range. The resulting bar pattern reveals the shape of the distribution — whether it is bell-shaped, skewed left, skewed right, or bimodal.

Right-click Sales in the Data pane and select Create → Bins. The Create Bins dialog opens. Set the bin size to 200 — this creates buckets of $0–$200, $200–$400, and so on. Click OK. A new field called Sales (bin) appears in Dimensions.

Drag Sales (bin) to the Columns shelf. Drag Number of Records (or CNT(Orders)) to the Rows shelf. Tableau renders a histogram — bars showing how many orders fall into each $200 Sales bucket.

Right-click the Sales (bin) field in the Data pane and select Edit to adjust the bin size. Smaller bins (e.g. $100) show finer detail. Larger bins (e.g. $500) show the overall shape more clearly. Choose a bin size that balances detail with readability.

Right-click the X axis → Edit Axis → uncheck Include Zero if the distribution starts well above zero. Go to Format → Borders and set the column divider to None — histograms read best without gaps between bars.

Histogram — Labelled Mockup

Sales Distribution — $200 bins · Count of Orders

Reading Distribution Shape

Right-skewed

Most values are small, a long tail extends to the right. Common in Sales and revenue data — most transactions are small, a few are very large. Mean is higher than median.

Left-skewed

Most values are large, a long tail extends to the left. Less common — seen in completion rates or satisfaction scores where most results cluster near the maximum.

Normal

Symmetrical bell shape — values cluster around the mean with equal tails on both sides. Common in measurement data such as height, weight, or manufacturing tolerances.

Bimodal

Two separate peaks — indicates two distinct sub-populations in the data that should probably be analysed separately rather than as a single group.

Box Plots — Five-Number Summaries

A box plot condenses an entire distribution into five values and displays them as a compact visual. The five numbers are the minimum, the 25th percentile (Q1), the median (Q2), the 75th percentile (Q3), and the maximum. The box spans Q1 to Q3 — called the interquartile range (IQR) — showing where the middle 50% of values sit. The whiskers extend to the minimum and maximum, and individual dots beyond the whiskers are outliers.

Box Plot Anatomy — Labelled Mockup

Sales Distribution by Region — Box Plot

Building a Box Plot in Tableau

Drag Region to the Columns shelf and Sales to the Rows shelf. Tableau shows a bar chart of SUM(Sales) by Region — one bar per Region.

On the Marks card, open the mark type dropdown and select Circle. Each Region now shows one dot at its aggregated SUM value — still not a box plot yet.

Drag Sub-Category to the Detail channel. Now each Region shows one dot per Sub-Category — the individual Sales values are disaggregated. The dots spread across a range, revealing the distribution.

In the Analytics pane, drag Box Plot onto the view. Tableau overlays the IQR box, median line, and whiskers on top of the individual dots. The full box plot is complete.

Histogram vs Box Plot — Choosing the Right Tool

Aspect	Histogram	Box Plot
Best for	Seeing the full shape of a distribution — peaks, skew, gaps	Comparing distributions across multiple groups side by side
Groups	One distribution at a time (or small multiples)	Multiple groups fit easily side by side
Outliers	Visible as isolated bars at the extremes	Shown explicitly as individual dots beyond the whiskers
Key output	Count of records per value range	Median, IQR, and outlier positions

📌 Teacher's Note

Most business analysts reach for averages and never look at the distribution behind them. This is a serious analytical blind spot. The Superstore Sales field is right-skewed — the vast majority of orders are small, but a handful of large orders push the mean well above the median. If you report average order value and use it to set sales targets, you are setting targets based on a number that most orders will never reach. A histogram makes this visible in seconds. For box plots, the most powerful use case is comparison: does the East region have a wider spread of order values than the West? Are outliers concentrated in one region? The box plot answers both questions in a single view that a bar chart of averages can never provide.

Practice Questions

1. Before building a histogram in Tableau, you need to create equal-width buckets from a Measure field. What are these buckets called and how do you create them?

2. You have disaggregated dots showing Sub-Category Sales by Region. What do you do to overlay the IQR box, median line, and whiskers?

3. A histogram of order values has a tall bar on the left and bars that shrink steadily toward the right. What distribution shape is this and what does it tell you?

Next up — Lesson 24: Treemaps — using nested rectangles to show part-to-whole relationships and proportional size at a glance.

← Previous Course Index Next →

Tableau Course