One study, too many windows
Excel for the data, SPSS for tests, Mplus for latent models, G*Power for sample size, R for plots, Word for the paper. Every handoff adds another file to track.
Connect data, methods, and results for psychology, education, and social science research.
Paste a spreadsheet, upload CSV, Excel, SPSS (.sav), or Stata (.dta), or try the example
Handle missingness and build scales
Assign variables and configure a model
Read visuals and export a report
Does this workflow look familiar?
Excel for the data, SPSS for tests, Mplus for latent models, G*Power for sample size, R for plots, Word for the paper. Every handoff adds another file to track.
Copy coefficients, align decimals, adjust tables, then repeat when a model changes. The last mile can take longer than the analysis.
Knowing which button to press is different from knowing why a method fits. Switching between a textbook and software breaks the train of thought.
Long data contexts and repeated corrections can consume tokens on metered AI services. A fluent answer still needs statistical checking.
Changing a prompt, model version or generated script can change the workflow. Save data, code, settings and random seeds—not just the chat.
Effect sizes, uncertainty and model assumptions all matter. A p-value alone cannot answer the research question.
How does an effect arise, and when does it change?
How consistent are your scale items?
Does your measurement structure fit the theory?
How do groups or experimental conditions differ?
How many participants does your study need?
How do you handle missing values and build scales?
The model diagram updates as you assign variables. Read coefficients and confidence intervals together to understand effect direction and uncertainty.
| Effect | Estimate | 95% confidence interval |
|---|---|---|
| Total effect | 0.33 | [0.18, 0.48] |
| Direct effect | 0.18 | [0.04, 0.32] |
| Indirect effect | 0.15 | [0.07, 0.26] |
Illustrative values showing the report format.
Export coefficients, intervals, and plots in Chinese or English.
Revisit configurations and results saved in this browser.
AI can assist with analysis, but it may skip key steps. PsyStat preserves the full analysis workflow so every result can be checked.
Calculations use the original data and statistical methods directly, rather than AI-generated results.
Every step is visible, from data preparation and parameter settings to the final results.
Analysis settings and past results are saved automatically for easy review and verification.
“After fitting a model, I used to spend hours organizing path coefficients, indirect effects, and confidence intervals. Now I no longer have to dig through stacks of output tables, and I can focus on the research questions that really matter.”
“The paper reporting feature surprised me most. I used to check numbers repeatedly and adjust the formatting. Now the results follow the conventions of psychology papers, making the Results section much easier to write.”
“The hardest part of teaching statistics is explaining why a method is appropriate, rather than which buttons to click. PsyStat makes the analysis process visual, helping students understand the statistical logic behind the results.”
| Feature | Excel | SPSS | Mplus | R | PsyStat |
|---|---|---|---|---|---|
| Data preparation | Manual | Supported | Data transformations | Coding required | Supported |
| Scale scoring & reliability | Manual | Supported | Model syntax | Coding required | Supported |
| Mediation & moderation | Cumbersome | PROCESS | Syntax required | Coding required | Auto-generated reports |
| CFA / SEM | — | Requires AMOS | Syntax required | Coding required | Visual model editor |
| Longitudinal RI-CLPM | — | Complex setup | Syntax required | Coding required | Visual model editor |
| Experimental analysis | Basic | Supported | Not a core feature | Coding required | Auto-generated reports |
| Power & sample size | Manual | Supported | Not a core feature | Coding required | Supported |
| Statistical plots | Basic | Tedious customization | Limited styling | Coding required | Auto-generated charts |
| Word reports | Manual | Manual formatting | Manual | Setup required | One-click Word export |
| Analysis history | Files | Files | Files | Scripts | Local history |
From data to a research report, in one place.
Loading plans…
Absolutely! We publish cross-checks of statistical results on the Verification page. Accurate calculations are only the first step. Whether you have chosen the right model depends on your research design. Statistics can do the calculations for you, but it cannot handle reviewers for you.
Absolutely! To hell with code. You do not need to learn Python, R, or Mplus first. Upload your data, choose your variables, set up your model, and leave the rest to PsyStat. We also provide the corresponding code in the guide on the right so you can verify the analysis. Understanding why you analyze the data this way matters more than knowing how to write the code.
Rest assured! PsyStat performs statistical calculations locally in your browser. Your research data is not uploaded to AI or synchronized to the cloud. Drafts and analysis records are also stored in your local browser. We do not collect any of your research data.
Of course! We have prepared example datasets. Load an example and you are ready to go.
Your analysis history is saved in the current browser storage and can be viewed under Analysis history. Unfinished drafts depend on the tab session and may be lost when you close it. Cloud sync is not available yet, so remember to export important results promptly. After all that work on your model, do not let an accidental browser-data cleanup wipe it out!