Accessible Instructional Data

• Give the table a concise title or caption that identifies its subject. State time period, population, geography, and units nearby.
• Mark column and, where needed, row headers as headers, not merely bold cells. Keep header text short and specific.
• Prefer a simple grid. Split a table with many merged cells into smaller tables when the relationship permits. When grouped headers are essential, verify that every data cell is associated with the right headers.
• Explain abbreviations, suppressed values, missing values, estimates, and data sources. Do not use a blank cell to mean several different things.
• Use color as a supplement. If red means an increase or risk, say so in text and include a label or value.
• For a scanned or image-only table, obtain source data or rebuild a real table. A single image description does not support cell-by-cell comparison.

Example: Instead of “Table 1” above an image of numbers, use “Average lab response time, by site and semester (minutes).” Give the table columns Site, Fall 2025, Spring 2026, and Change (minutes), and define any unavailable value in a note.

• Put a descriptive title or short orientation at the start of each sheet. Name sheets for their contents.
• Use one clear header row for each data table. Keep one variable per column, consistent units and formats, and a stable meaning for each row.
• Avoid blank rows and columns inside a data range, merged cells used as headings, and information conveyed only through cell fill color.
• Provide a data dictionary for variable names, units, categories, missing-value codes, and source or collection date.
• Check the workbook with its accessibility checker, then navigate the key sheet with a keyboard and screen reader when possible. Review charts separately.

Microsoft’s Excel accessibility guidance recommends orienting readers at the start of a sheet and using meaningful headers. For web tables, use the W3C tables tutorial.

• Use a meaningful chart title, axis labels, units, series names, and visible data labels where useful.
• Write a short text alternative that identifies the chart and its purpose. Put a longer explanation adjacent to a complex chart, covering the pattern, comparisons, and meaningful exceptions.
• Provide underlying values in a table or accessible download when precision matters.
• Use patterns, direct labels, or symbols as well as color. Check text and graphical contrast.
• For interactive charts, ensure every filter, legend control, tooltip, and data point needed for the task is reachable and understandable by keyboard. Offer an accessible table when the chart cannot expose the needed values.

The W3C complex-image tutorial explains the short-plus-long-description pattern. Microsoft’s chart guidance covers descriptive titles, axes, and labels.

Treat the whole task as the unit of review: find a dataset, understand filters, set them, read results, export data, and recover from errors. Check:

• A page title, headings, and a short orientation explaining what the dashboard contains.
• Labels for all filters, search fields, date ranges, and buttons. State when applying a filter changes results.
• Keyboard access and visible focus throughout. Focus should not disappear or move unpredictably after refresh.
• Results announced or discoverable after a filter changes; no silent updates that leave users unsure what happened.
• A navigable data table or download with the same essential information as each visual.
• Definitions for measures, denominators, source dates, suppression rules, and uncertainty.
• Usable behavior at zoom and on narrow screens, plus readable contrast and non-color cues.