GraphHelix analyzes your data structure and research question, then recommends the right test — with assumption checks, effect sizes, and publication-ready output.
Built for researchers who collect data but don't write code.
Already have access? Sign in
Describe your research question — "I'm comparing three treatment groups on a continuous outcome" — and GraphHelix recommends the right test with a clear explanation of why.
Before running any test, GraphHelix checks normality (Shapiro-Wilk), variance homogeneity (Levene's test), and other assumptions your data must meet. If assumptions are violated, it suggests a non-parametric alternative — one click to switch.
Every test reports the full set: test statistic, degrees of freedom, p-value, effect size with magnitude label, and 95% confidence interval. Formatted in APA 7th edition notation — t(28) = 3.45, p = .002, d = 0.82 [medium] — ready to copy into your manuscript.
After every test, GraphHelix explains what the results mean in plain English. Not sure about something? Ask follow-up questions in the AI chat — it has full context of your data and analysis.
Powerful tools wrapped in a simple interface so you can focus on your research, not your software.
Parametric and non-parametric tests, Bayesian analysis, survival analysis, regression, PCA, mediation and moderation — each with assumption checks, effect sizes, and APA-formatted output.
One-click copy of APA 7th edition result strings: t(28) = 3.45, p = .002, d = 0.82. Every test reports the complete set — test statistic, degrees of freedom, p-value, effect size, and confidence interval.
Export charts as PNG or SVG at 150, 300, or 600 DPI. Journal presets for Nature, Science, PLOS ONE, JAMA, and Cell automatically size your figures to each publication's requirements.
Calculate the sample size you need before collecting data. Supports t-tests, ANOVA, chi-square, and correlation with interactive power curves and Cohen's effect size benchmarks. Try the free calculator
Plan your study before collecting data. Describe your research design to the AI — it recommends statistical tests, calculates sample size, and helps draft a pre-registration document.
Drag and drop CSV, Excel, SPSS (.sav), or Stata (.dta) files. Variable labels and value labels are preserved automatically. Switching from SPSS or Stata? Your existing files work here.
Test statistic, degrees of freedom, p-value, effect size, confidence interval, and APA-formatted output — for every test.
Every test includes automatic assumption checking. When assumptions are violated, GraphHelix suggests the appropriate alternative — one click to switch.
Need something custom? Write Python scripts with numpy, pandas, scipy, and statsmodels in a sandboxed environment.
You're in control. The AI explains its reasoning so you can make the final call.
GraphHelix reads your dataset's column types, group structure, and variable relationships. It identifies whether your outcome is continuous or categorical, how many groups you're comparing, and whether observations are paired or independent.
Before recommending a test, GraphHelix runs Shapiro-Wilk normality tests and Levene's test for equal variances on your data. If assumptions are violated, it recommends robust alternatives — and explains why.
The recommendation includes the specific test, why it fits your data, which assumptions were checked, and what alternatives exist. You see the reasoning, not just the answer — so you can justify your choice to reviewers.
GraphHelix suggests — you decide. Run the recommended test, ask follow-up questions, or choose a different approach. The AI augments your judgment; it doesn't replace it.
Go from raw data to actionable results in minutes, not hours.
Import a CSV, Excel, SPSS, or Stata file — GraphHelix detects column types and data structure automatically. Or try example data to explore without uploading anything.
Describe your research question and variables. GraphHelix evaluates your data's distribution, checks assumptions, and recommends the most appropriate test — with a clear explanation of why. If assumptions are violated, it suggests robust alternatives.
View the complete result: test statistic, p-value, effect size, confidence interval, and APA-formatted string ready to paste into your manuscript. Export charts at journal-quality resolution.
All data is transmitted over TLS and stored in encrypted managed databases. API keys are encrypted with AES-256-GCM.
Research datasets are not used to train AI models. Your data is processed for your analysis only.
Projects are private by default. Sharing is opt-in with expiring links.
Full privacy policy coming soon.
Join the beta waitlist and be the first to try GraphHelix.