The six-step workflow
Start with a prepared spreadsheet. Review each important choice along the way.
Upload your dataset
Choose a CSV or XLSX file. For Excel, your workbook must contain exactly one visible worksheet.
Your starting point: one table, clear column names.
Get to know your data
Review the data overview and descriptive statistics. Check missing values, duplicate rows and distributions before moving into an analysis.
Understand what is in your spreadsheet.
Confirm types and exclusions
Review whether variables are numeric, unordered categories, ordered categories or dates. Confirm which columns to exclude, and check what coded values mean.
A number such as “1” may be a category, not a quantity.
Ask in natural language
Describe what you want to understand and name the relevant variables. Confirm important variable roles and analysis choices when prompted.
Example question
“Does the average exam score differ between teaching groups?”
Explore results and explanations
Statistical engines compute a supported analysis. Review the results and relevant charts, then read the AI explanation of the findings, uncertainty and limitations.
Check the method and assumptions alongside the answer.
Export your PDF report
Save your results, charts and explanations in a PDF report to review or share. Export the reports you need before leaving your temporary workspace.
Keep a copy with your original dataset.
Statistics first, AI second
Dedicated statistical engines calculate your results. AI helps interpret your question and explains those computed results in accessible language.
You stay in control
Your knowledge of the dataset matters. Important variable roles and analysis choices require your confirmation before proceeding.
Check the proposed outcome, comparison groups and other relevant variables when asked. Make sure their coding and roles match your research question.
What you receive
- A data overview: descriptive statistics and data quality checks.
- Computed results: structured statistical output for the supported analysis.
- Relevant charts: visualisations suited to the analysis.
- AI explanations: readable findings with context on uncertainty, assumptions and limitations.
- A PDF report: an export you can keep and share.
Uploads and descriptive statistics are available for free. Full analyses, AI explanations and PDF export require credits; one completed analysis uses one credit. If an analysis cannot be delivered, the credit is returned automatically.
Before you upload
- Prepare a CSV or XLSX file; Excel must have exactly one visible worksheet.
- Use clear, unique column names and include only relevant rows and columns.
- Understand your variable coding, including category labels and their order.
- Check missing values and obvious data entry errors.
- Upload only data you have permission to process. Remove or anonymise unnecessary personal information, including in free-text cells, column names and questions.
Available methods
See the current list of supported methods.
Limitations
Statsopia supports a defined set of analyses. Some questions or study designs may need a method that is not available in Beta.
Results depend on data quality, variable coding and an appropriate analysis. AI can misunderstand a question or explain a result incorrectly. Review the method, assumptions and findings before relying on them; expert judgement remains important for consequential decisions.
Privacy and temporary storage
Your dataset is processed to provide the analysis. AI receives your question and relevant data context or computed results to interpret the request and explain findings.
Dataset access expires two hours after storage. This access window is not a guarantee that every stored copy is erased at that moment.
Refreshing the page ends your current workspace continuity. Signing out or replacing the dataset clears the active workspace. Keep your source file and export any reports you need.
The Privacy Policy describes data handling in full.