Use Cases Overview Group Comparison & ANOVA ↳ Two-Group Comparison ↳ One-Way ANOVA ↳ Two-Way ANOVA ↳ Repeated Measures Regression & Curve Fitting ↳ Linear Regression ↳ Logistic Regression & AUC ↳ Dose-Response & IC50 ↳ Mixed Models (GLMM) Survival Analysis ↳ Kaplan-Meier Workflow ↳ Cox Regression Workflow Normality & Power ↳ Pre-Study Power Calculation ↳ Normality Testing Categorical & Association ↳ Chi-Square Test ↳ Fisher's Exact Test ↳ Odds Ratio & Risk Life Sciences Applications ↳ Flow Cytometry ↳ Gene Expression / Omics ↳ Animal Study Workflow ↳ Clinical Trial Endpoints

Group Comparison & ANOVA

Comparing two or more groups on a continuous outcome is the most common task in experimental research. Licklider walks you through the full path — from normality check and variance homogeneity to test selection, effect size, and journal-ready figure — while enforcing every mandatory disclosure.

Verified execution paths (parametric-first): Independent t-test · Welch's t-test · One-way ANOVA · Two-way ANOVA · Post-hoc (Tukey, Holm, Bonferroni, etc.) · Effect size (Cohen's d, η², ω²). Guidance-only in the current product: Mann–Whitney / Kruskal–Wallis / Wilcoxon-style rank tests — see Non-Parametric Alternatives. Repeated-measures ANOVA / LMM: coverage varies; check method pages and Known Limitations.

When to Use This Section

Use Group Comparison & ANOVA workflows when:

  • You have a continuous outcome variable and one or more categorical grouping factors
  • You want to test whether group means (or distributions) differ significantly
  • You need to report effect size, CI, and post-hoc results alongside significance annotations
  • Your design includes repeated measurements, paired observations, or factorial combinations

Available Scenarios

Test Selection Guide

Licklider's automatic selection and engine execution prioritize parametric routes (Welch, ANOVA, post-hoc). The UI may surface rank-based method labels when assumptions suggest them, but Mann–Whitney, Wilcoxon, Kruskal–Wallis, and Friedman are not verified end-to-end through the public stats engine — see Non-Parametric Alternatives.

  • 2 groups, independent, normal: Welch's t-test (default over Student's t)
  • 2 groups, independent, non-normal: Selection may recommend a rank-based label; execution remains parametric-first — use the guidance page to plan outside-Licklider analysis if needed
  • 2 groups, paired: Paired t-test when a subject/block ID column is specified (verified path)
  • 3+ groups, independent: One-way ANOVA → Holm or Tukey post-hoc (verified)
  • 3+ groups, non-normal / unequal n: Review Non-Parametric Alternatives; no dedicated rank-based omnibus path is verified in-product
  • 2 factors: Two-way ANOVA with interaction term (where supported)
  • Repeated measures: See method pages; not all RM/LMM paths are fully verified — Known Limitations