Statistical data visualization methods
Data visualization options
1. Histogram
The histogram below illustrates the distribution of monthly sales revenue for a retail store over the past year. The data is grouped into revenue ranges (bins) of USD 10,000 increments, with the frequency represented by the height of the bars. The x-axis represents the ‘Revenue Range in USD’, while the y-axis represents the ‘Number of Months’. The histogram shows how often monthly revenue fell in each range, which reveals the most common range, the spread, and any unusually high or low months. It does not show which months those were or the order they came in, since binning discards time order. Use a line chart to see trends over time.
When to use histograms
Histograms are best used when analyzing:
- The distribution of a dataset to understand its shape (e.g., normal, skewed, bimodal).
- The frequency of values within specific ranges.
- The presence of outliers or gaps in the data.
- How much the values vary (spread).
2. Box plots
The box plot below illustrates the monthly transaction amounts of an e-commerce business over a year. The x-axis represents different months, while the y-axis represents the transaction values in dollars. The box represents the interquartile range (middle 50% of transactions), the line inside the box marks the median transaction value, and the whiskers typically extend to the most extreme values within 1.5 × IQR of the box; values beyond that are plotted individually as outliers.
A few interpretations can be obtained from the box plot above:
- December’s high transaction values are shown by a higher median and upper quartile, indicating increased consumer spending due to holiday shopping (e.g., Black Friday, Christmas).
- Wide interquartile ranges (IQRs) in June, October, and November suggest high variability in transaction amounts, often driven by seasonal promotions and major shopping events.
- Lower median and compact IQRs in January and May indicate lower and more consistent transactions, likely due to post-holiday budget constraints and fewer shopping incentives.
- Long whiskers in months like October and November reveal a broad range of transaction values, meaning customers spent at very different levels, while shorter whiskers in months like January suggest more stable spending behavior.
When to use box plots
Box plots are best used when analyzing:
- The distribution of a dataset and identifying its central tendency.
- The presence of outliers that may affect analysis.
- The comparison of distributions across multiple categories or groups.
- The overall spread and variability of data.
3. Scatter plot
The scatter plot below visualizes the relationship between marketing spending and monthly sales revenue for an online retail store over the past year. Each point represents one month’s data, with the x-axis showing the marketing spend (in dollars) and the y-axis showing the corresponding sales revenue. By analyzing this scatter plot, the company can determine whether higher marketing spend is associated with higher sales and identify whether any months show unusual deviations from the expected trend.
The scatter plot shows a clear upward trend: months with higher marketing spend tend to have higher revenue, a positive, roughly linear association. A scatter plot shows association, not causation - a third factor such as seasonality could be driving both variables, so a causal claim needs further analysis.
Some months deviate from this trend, meaning the correlation isn’t perfect; other factors beyond marketing also affect sales performance.
When to use scatter plots
Scatter plots are best used when analyzing:
- The relationship between two numerical variables (e.g., sales revenue vs. advertising expenditure).
- The presence of positive, negative, or no correlation between variables.
- Trends, clusters, or outliers in data distribution.
- See whether higher values of one variable tend to go with higher or lower values of the other (positive or negative association).
4. Dot plot
The dot plot below illustrates the monthly inventory levels of a retail store over the past two years. Each of the twenty-four (24) dots represents the inventory count at the end of a given month, plotted along the x-axis, which shows the number of inventory units. By analyzing this visualization, store managers can identify common inventory levels, fluctuations, and potential stock management issues that may impact operations and profitability.
When to use dot plots
Dot plots are best used when:
- Displaying small to moderately sized datasets where individual values need to be seen.
- Identifying clusters, gaps, and outliers in data.
- Analyzing frequency of occurrence in a numerical dataset.
Choosing among these charts
| Chart | Best for | Key limitation |
|---|---|---|
| Histogram | Shape of one continuous variable | Bin width choice can change the apparent shape, and time order is lost |
| Box plot | Comparing spread or outliers across groups | Hides modality and sample size |
| Scatter plot | Relationship between two variables | Shows association, not cause; axis scaling can exaggerate the trend |
| Dot plot | Small datasets where individual values matter | Gets cluttered with large datasets |
Watch the axes on any of these charts: a distorted scale is one of the easiest ways to mislead a reader. Check that any axis where bar height carries the value, such as a histogram’s frequency axis, starts at zero; scatter, box, and dot plots can use a non-zero axis as long as it is clearly labeled. Keep scales consistent across charts you’re comparing.