Our Mission
In biomedical research and clinical practice, investigators frequently require rapid statistical calculations—such as verifying sample sizes for grant submissions, checking $p$-values during manuscript peer review, or computing odds ratios for epidemiological surveillance.
Commercial statistical packages (such as SPSS, SAS, or Stata) can be cost-prohibitive for early-career investigators and global health teams. Meanwhile, command-line environments like R require coding setups that may not be readily available on mobile devices or in clinical environments.
LabStats was built to bridge this gap: providing instantaneous, rigorously validated biostatistical calculators that run directly in any modern browser without software installations, telemetry, or paywalls.
Computational Rigor & Benchmarking
Scientific integrity depends on computational accuracy. Our mathematical engine (stats-core.js) is built on peer-reviewed algorithms:
- Inverse Standard Normal CDF: Implemented using Peter J. Acklam's rational approximation algorithm, providing a maximum relative error of $< 1.15 \times 10^{-9}$.
- Log-Gamma Function: Computed using the Lanczos approximation ($g=7, n=9$), maintaining approximately 15 decimal digits of precision across the positive real line.
- Incomplete Beta & Gamma Functions: Evaluated using continued fractions via modified Lentz methods and Taylor series expansions, ensuring robust convergence even for extreme degrees of freedom.
All algorithms are cross-validated against the gold-standard statistical libraries in R (version 4.x) and Python SciPy (1.13+) before deployment.
Commitment to Data Privacy
Healthcare institutions strictly forbid uploading patient counts or unpublished clinical trial data to unverified third-party cloud servers. To ensure compliance with HIPAA, GDPR, and institutional review boards (IRBs), all calculations on LabStats are performed 100% locally in your client's web browser. No medical or numerical data is ever transmitted to our servers.