Use line charts to foreground x–y interactions

Choose a line chart when the main message is the x–y relationship and the legend variable should read as the moderator. For example, use connected lines instead of grouped bars for three-variable data when you want viewers to say how y changes across x for each legend-coded group.

Connected lines make the x-axis relationship the dominant visual chunk, so viewers tend to read change across x first and treat the legend variable as the condition on that relationship.

Mechanism: Good continuity groups points into lines, which makes the x–y relationship visually salient and encourages interaction descriptions framed around the x-axis variable.

Evidence: In written descriptions of multivariate graphs, viewers described x–y interactions more often from line graphs than from bar graphs, and this x–y emphasis was strongest when the content was unfamiliar. (Shah & Freedman, 2011)

Notes: The line chart created a stronger directional bias than the bar chart.

• User Goal: The reader should state the x–y interaction as the main point.
• Task: Open-ended explanation of what matters most in a three-variable graph.
• Data: Two ordered independent variables and one quantitative dependent variable.
• Chart Setting: One variable is on the x-axis and the other grouping variable is shown in the legend as separate lines.
• Audience: Topic familiarity may be low; the line-chart emphasis on x–y interactions was strongest for unfamiliar content.
• Success Criterion: Readers spontaneously describe how y changes across x, qualified by the legend-coded groups.

Break it when: The intended message is a comparison among legend categories within each x-axis group, or a main effect that ignores one variable. Why: Line charts steered viewers away from those readings and toward x–y interaction descriptions.

Sacrifice: You give up emphasis on within-group legend-category comparisons and on main-effect summaries.
Risk: Readers may overlook alternative summaries of the same data because the connected lines strongly cue one interpretation.
Mitigation: If those alternative summaries are the message, switch to a grouped bar chart instead of trying to force them out of the line chart.

Mistake: Using a line chart when the key message is the legend-variable comparison inside each x-axis group. Why it fails: The line chart encourages readers to describe change across x instead of the intended grouped comparison.

Failure Sign: Reviewers lead with a different summary than the intended x–y interaction.
Quick Check: Show matched line and bar versions and ask, “What is the main point?” Choose the line chart only if the x–y interaction is mentioned more readily from the line version.
Stronger Test: Run the same prompt with viewers unfamiliar with the topic, since reliance on the line chart’s x–y cue was strongest there.

• Replot the grouped bars as connected lines while keeping the same variables on the x-axis and in the legend.
• Retest with an open-ended “main point” prompt and keep the line chart only if readers now lead with the intended x–y interaction.
• If readers still need to notice grouped legend-category comparisons or main effects instead, replace the line chart with a grouped bar chart.