Skip to main content
Verify the figures and statements you intend to deliver. Start in Charts, where the plan folder holds each planned question’s results, then check the deck built from your folder in Slides. Use Data Analyst for questions that need supporting evidence.

Before you start

Open the intended study and output version. Note the weighting, response filters, and collection stage. Use the analysis plan to identify the questions and comparisons that the outputs should cover.

Review the deck

Verify the numbers before the story. In Analysis → Charts, open the plan folder and step through it question by question: the outputs of each question split, reading the raw bases before the coloured cells and letters. Gather the charts you will rely on in a folder of your own. Then, in Analysis → Slides, choose the deck conversation built from that folder and use the filmstrip to move between slides. Regenerating into the plan folder deletes and rebuilds its charts, including any you filed there, but does not rewrite saved decks. Take one important claim at a time:
  1. Identify the source question and comparison group.
  2. Open the corresponding chart or table in Charts.
  3. Check the included raw base, weighting, and filters.
  4. Compare the displayed result with the slide’s text.
  5. Read supporting responses when the claim interprets open answers.
Saved text can describe an earlier collection stage while chart values reflect current responses. Regenerate from the plan, then generate slides from the folder again or request a targeted revision, and review again before exporting.

Inspect a chart or table

Use a saved chart’s controls to inspect its question, breakdown, filters, and weighting. Save the chart when the comparison must be repeated; its name follows the question, theme and breakdown. A filter changes the displayed population; it does not rewrite slide prose or earlier assistant messages. In crosstabs, distinguish raw and weighted bases. Compare eligible columns and inspect the selected tests. Selected tables are generated by default. Only when Only keep tables with a significant difference is enabled does the plan remove comparisons without a qualifying significant cell on a raw column base of at least 30. Review that option, question eligibility and selected comparisons before diagnosing a missing table. An untested or unmarked difference is not evidence that groups are identical. See Interpret significance. Click the chart segment to open its respondents directly. The list follows that segment’s filters and grouping and provides pagination and Show all respondent answers for context. Download Excel from the panel for the listed subset. Switching charts or measures closes the panel; reopen the segment you intend to inspect.

Review themes and personas

Themes summarize open answers. An answer can belong to more than one theme, so theme counts can exceed the number of respondents and percentages need not total 100%. Use Charts or Data Analyst to investigate a theme and its supporting responses. Read a selection of included answers, check contradictory evidence, and distinguish the researcher’s interpretation from the participant’s wording. Where available, personas provide another grouping for charts. Review the responses supporting a persona before treating the grouping as a finding.

Open cited evidence

Select a response citation in the assistant’s answer. Read the complete source, including its recording or transcript when available. Show all respondent answers provides the surrounding answers for that participant. Check that the answer concerns the intended product, time period, or condition. Do not treat a selected quotation as proof of prevalence; use the corresponding counts and base.

Ask a focused follow-up

Attach the saved chart or table to Data Analyst and name the cell, group, or question you need explained. Ask for the supporting bases and citations. The assistant can read the computed crosstab; verify its interpretation against the displayed table.

Check the result

Each retained figure and quotation has a checked source. The chart population matches the wording, themes have supporting answers, and the chosen deck reflects the analysis version you intend to deliver.