Use stacked bar charts instead of stacked line charts for negative correlation judgments

Use a stacked bar chart instead of a stacked line chart when viewers need to judge the strength of a negative correlation. For example, if two quantitative series move in opposite directions and you are choosing between these two stacked forms, use stacked bars rather than stacked lines.

Among the stacked variants, stacked bars gave the strongest negative-correlation judgments. Stacked lines were a weaker option for the same task.

Mechanism: The stacked bar form made differences in negative relationship strength easier to discriminate than the stacked line form, which supported less precise judgments.

Evidence: For negatively correlated data, stacked bar charts significantly outperformed stacked line charts in the paper’s correlation-judgment experiment and Weber-model analysis (Harrison et al., 2014).

Notes: The paper reports participant comments suggesting that viewers attended to different visual cues in stacked lines than in stacked bars.

• User Goal: Let viewers decide which negative relationship is stronger.
• Task: Compare association strength between two quantitative series.
• Data: Two quantitative series with negative correlation.
• Chart Setting: A stacked presentation is desired and the choice is between stacked bars and stacked lines.
• Success Criterion: Viewers can more reliably distinguish stronger from weaker negative correlations.

Break it when: The relationship to be judged is positive rather than negative. Why: Positive stacked-line and positive stacked-bar conditions frequently hit the study’s chance boundary, so this contrast was not supported for positive correlations.

Sacrifice: This rule only resolves the negative-correlation case. Risk: Applying it blindly to positive correlations can claim a benefit the study did not establish. Mitigation: Choose between stacked forms separately for positive and negative relationships.

Mistake: Keep the stacked line chart because it appears visually similar to the stacked bar alternative. Why it fails: The study found that these similar-looking stacked forms did not provide the same correlation-judgment precision.

Failure Sign: Viewers disagree about which negative stacked-line relationship is stronger. Quick Check: Show the same negative relationship as stacked bars and as stacked lines, then ask which version makes the stronger correlation easier to see. Stronger Test: Run repeated paired comparisons on both chart types and keep the one with more consistent correct choices.

• Replace the stacked line chart with a stacked bar chart for the negative-correlation view.
• Compare both versions on the same negative dataset before standardizing a stacked form.
• Separate positive and negative relationship cases instead of reusing one stacked chart choice everywhere.