Yes. All data is encrypted in transit (TLS) and at rest. Your research data is never used to train AI models. Projects are private by default, and sharing is opt-in with expiring links.

Frequently asked questions

Q: Does it work with SPSS files?

Yes. GraphHelix imports .sav files with full support for variable labels and value labels. Your existing SPSS files work without conversion.

Does GraphHelix work for non-parametric tests?

Yes. GraphHelix supports Mann-Whitney U, Wilcoxon signed-rank, Kruskal-Wallis H, Friedman test, and Spearman correlation. When a parametric test's assumptions are violated (e.g., Shapiro-Wilk rejects normality), GraphHelix suggests the appropriate non-parametric alternative and lets you switch in one click.

Can I export APA-formatted results?

Yes. Every statistical test produces an APA 7th edition formatted result string — for example, t(48) = 3.45, p = .001, d = 0.97 — that you can copy with one click. The string includes the test statistic, degrees of freedom, p-value (no leading zero, per APA convention), effect size with magnitude label, and 95% confidence interval. You can also generate a full APA Methods & Results section from the "Export & History" dropdown.

Is my data private?

Yes. All data is encrypted in transit (TLS) and at rest. API keys are encrypted with AES-256-GCM. Your research data is never used to train AI models — it is processed for your analysis only. Projects are private by default, and sharing is opt-in with expiring links. Full privacy policy coming soon.

Does it work with SPSS files?

Yes. GraphHelix imports .sav files directly with full support for variable labels, value labels, and metadata. Your existing SPSS files work without any conversion. We also support Stata .dta files with variable label preservation. See Data format requirements for details.

Can I use it for free?

GraphHelix is currently in beta. Join the waitlist to request access — beta testers have access to all features at no cost. Pricing details will be announced before general availability.

What statistical tests are supported?

GraphHelix supports 30+ tests including t-tests (unpaired, paired, one-sample), ANOVA (one-way, two-way, repeated measures), regression (linear, logistic), correlation (Pearson, Spearman), non-parametric tests (Mann-Whitney, Wilcoxon, Kruskal-Wallis, Friedman), Bayesian tests (t-test, ANOVA, linear regression), survival analysis (Kaplan-Meier, Cox proportional hazards), and advanced methods (PCA, EFA, mediation, moderation, LMM). See the full list.

Does GraphHelix check statistical assumptions automatically?

Yes. Before every parametric test, GraphHelix runs Shapiro-Wilk normality tests (per group), Levene's test for equal variances, and Mauchly's sphericity test (for repeated measures). If assumptions are violated, it recommends the appropriate alternative test and explains why.

Can I write my own analysis scripts?

Yes. GraphHelix includes a sandboxed Python environment with numpy, pandas, scipy, and statsmodels. You can write custom analysis scripts using a CodeMirror editor with Python syntax highlighting. Scripts run in an isolated subprocess with resource limits for safety.

Does GraphHelix support Bayesian analysis?

Yes. GraphHelix offers Bayesian t-tests, Bayesian ANOVA, and Bayesian linear regression. Each reports a Bayes factor (BF₁₀) with a 5-step Jeffreys scale indicator, plus 95% credible intervals. These complement frequentist tests — you can run both on the same data to compare approaches.

How do I export figures for journal submission?

Charts can be exported as PNG (at 150, 300, or 600 DPI) or SVG. Journal presets for Nature, Science, PLOS ONE, JAMA, and Cell automatically size your figures to each publication's requirements. A filename preview is shown before download.

Have a question not answered here? Email us at hello@graphhelix.ai — we read every message and respond within a day.