Choosing the Right Picture for the Data
Different data situations call for different visualizations, and choosing the wrong one can obscure or actively mislead. A histogram shows the distribution of a single CONTINUOUS variable, grouping values into bins to reveal its overall shape (normal, skewed, bimodal). A bar chart compares values across distinct CATEGORIES (with gaps between bars, unlike a histogram's touching bars, signaling that the categories are not continuous). A scatterplot shows the relationship between two CONTINUOUS variables, one point per case, revealing patterns a correlation coefficient alone can only summarize numerically. A box plot (or box-and-whisker plot) visually summarizes a data set's five-number summary (minimum, first quartile, median, third quartile, maximum) at a glance, making outliers easy to spot. A line graph shows how a variable changes over ordered, typically time-based, intervals.
| Chart | Best for |
|---|---|
| Histogram | Distribution shape of one continuous variable |
| Bar chart | Comparing values across categories |
| Scatterplot | Relationship between two continuous variables |
| Box plot | Five-number summary and outliers at a glance |
| Line graph | Trends over time |
Visualizations can also actively MISLEAD, even using accurate underlying data. A truncated y-axis (starting at a value other than zero) can make a small difference look dramatic. Inconsistent scaling, cherry-picked time windows, and 3D effects that distort area/volume perception are further common tricks (intentional or not) that visually exaggerate a pattern beyond what the actual numbers support.
Common pitfall: using a bar chart (with a truncated y-axis) to compare two very similar values, making a genuinely small difference look enormous. Always check where a chart's axis actually starts — a bar chart's height comparisons are only honestly interpretable when the axis begins at zero.