Show the Pattern Without Hiding the Exact Values

A table chart combines exact values with a visual comparison. Use a table when readers need to look up numbers, a chart when they need to spot a pattern, and both when they need to check the numbers behind that pattern. For district service requests, show resolution rates as bars while keeping requests received and resolved in the same table.

Choose rates or resolved counts; compare the ranking and the workload behind it.

The term also describes tables with bars or colour inside cells, or interactive table visualizations. For example, Google Charts’ Table visualization displays rows and columns that readers can sort and page through; it is not a conventional graph. The example here uses in-cell bars to keep the comparison beside its exact values.

Choose the Display That Answers the Reader’s Task

Reader’s task	Recommended display
Look up an exact value	Table with clear labels and units
Compare several measures for one category	Table with adjacent measure columns
Spot the largest or smallest category	Bar chart with an accompanying table
Follow changes over time	Line chart with a dated source table
Explore many records	Filterable table with a focused chart or summary

This distinction matches Government Analysis Function guidance: tables support comparisons of values and summary statistics, while charts show patterns, trends and relationships.

Using both does not mean plotting every column. Give the chart one comparison to communicate; let the table supply the detail. In the district example, a rate comparison answers a different question from a count comparison, so the display must name the selected measure.

Central Leads on Rate; North Leads on Resolved Requests

This is a fictional sample dataset, not a report of actual district performance. Each row represents one district’s requests received during September 2026. The resolved count includes only requests from that same group resolved by month-end—not older requests cleared during September.

District	Received	Resolved	Rate (%)
North	120	96	80.0
Central	40	36	90.0
South	80	60	75.0

Resolution rate (%) = requests resolved ÷ requests received × 100.

North resolved the most requests from this group: 96. Central resolved the highest share: 90%. These are different findings, and neither should be substituted for the other.

For a rate comparison, order the bars Central, North, South and label the value axis “Resolution rate (%)”. Keep the counts in the accompanying table so readers can see that Central’s percentage comes from a smaller workload.

A suitable chart title is “Central had the highest resolution rate.” The subtitle should specify “Requests received in September 2026; status at month-end; fictional data.” A standalone chart needs this context even if the source table appears elsewhere on the page.

Neither the percentage nor the count alone establishes service quality. This example contains no information about request difficulty or staffing.

Keep the Source Fields and Calculations Consistent

The publication file should have one header row, no merged headings and no decorative blank rows. Each column should hold one field. For this example, the source fields are district, requests_received, requests_resolved and resolution_rate_pct; the published table can use shorter, readable headings.

Keep numeric fields numeric. Store the rate as 80.0, meaning 80%, and document the percentage unit in the header or description. Do not mix 80.0, 80% and 0.8 in the same field. That inconsistency can make a chart or imported table interpret comparable records differently.

Keep totals separate from district records. An “All districts” row should not become another bar in a district comparison. The overall rate is 192 ÷ 240 × 100 = 80.0%, not the simple average of the three district percentages. Calculate it from the combined resolved and received counts.

Define missing values before publication. A blank rate must not silently become zero. If no requests were received, the calculation has a zero denominator; document the rate as not applicable rather than showing a measured 0%.

Keep full calculation precision and format the displayed values consistently. The government table guidance recommends consistent precision within each column and notes that rounding can make displayed components differ from totals. Here, all displayed rates use one decimal place.

For a broader file-cleaning workflow, see how to prepare and publish a data table.

Check the Public Page Against the Source Table

As of October 2026, TablePage accepts CSV, TSV, XLSX and XLS uploads and creates public dataset pages with a filterable table and automatically generated charts. Upload only a publication-ready copy: remove personal information, confidential fields and anything else that should not be public.

Before sharing the page, compare its chart with the source records. A resolution-rate chart must use the same cohort as the table, not all requests resolved during the month. Check that a stored value of 80.0 is interpreted as 80%, rather than multiplied by 100 again.

Make the time period, denominator and fictional-data explanation easy to find. Readers should be able to reach the exact counts and rates without relying on a screenshot. For this sample, the page should show Central’s higher rate immediately while keeping its received count of 40 visible beside North’s 120 and South’s 80.

Give the web table a descriptive caption. W3C guidance explains that captions help readers, including screen-reader users, identify a table; a structural summary is usually needed only for complex tables. See writing accessible table captions and summaries for examples.

The finished display should preserve both findings: Central leads on resolution rate, while North leads on resolved count. The table supplies the evidence; the visual comparison makes the selected finding easier to see.