Add a human-readable data table when the chart alone does not convey all relevant information

Add a human-readable data table alongside the chart when readers need information that the chart title, summary, context, or annotations do not already fully convey. For example, pair a line chart with a table listing the datapoints, and keep the chart as well because the table does not replace the chart’s narrative or structural relationships.

A chart and a table expose different parts of the same information. The chart carries the higher-level pattern and relationships, while the table exposes the underlying values in a form that can be read directly.

Mechanism: A companion table gives readers access to low-level data without forcing the chart to do all exact-value work, and the chart remains available for the broader narrative that a table does not provide.

Evidence: Chartability marks the absence of a table as a critical accessibility failure and illustrates that a chart-table pairing can provide low-level datapoint access while preserving the chart’s higher-level reading; the linked NCAM guidance also recommends including data tables where appropriate for complex visuals. (Elavsky et al., 2022; Effective Practices for Description of Science Content within Digital Talking Books, n.d.)

• User Goal: Access the underlying values as well as the chart’s overall takeaway.
• Task: Read individual datapoints or inspect the data the chart is based on.
• Data: The chart contains relevant information not fully conveyed by the existing title, summary, context, or annotations.
• Chart Setting: A chart is already present and can be paired with a companion table.
• Audience: Readers who use assistive technologies or otherwise prefer a different information flow.
• Success Criterion: Readers can access both the high-level chart narrative and the low-level data values.

Break it when: The chart title, summary, context, or annotations already convey all relevant information contained in the chart. Why: The source explicitly allows the table to be excluded in that case.

Sacrifice: A companion table is an added representation, not a shortcut that removes the need to keep the chart accessible. Risk: Treating the table as a full substitute can remove the narrative and structural relationships that the chart communicates. Mitigation: Keep both the chart and the table available, using the table for low-level values and the chart for higher-level structure.

Mistake: Replacing the chart with a table and treating the two as equivalent experiences. Why it fails: A table may expose the data, but it does not provide an equivalent narrative and may not preserve relationships or interactions that are non-tabular.

Failure Sign: The chart shows values or relationships, but no companion human-readable table is available. Quick Check: Ask whether a reader can get all relevant information from the chart title, summary, context, or annotations alone. Stronger Test: Verify that the chart has a companion table containing the data it is based on and that the chart remains available for the higher-level reading.

• Add a companion human-readable table that contains the data the chart is based on.
• Keep the chart and the table together so readers can use the chart for the overall pattern and the table for exact values.
• If a table is unnecessary because the chart already communicates everything relevant, revise the title, summary, context, or annotations until they convey all relevant information contained in the chart.