Regression and Modeling

Use this overview to decide when a regression model is the right tool, choose the leaf page that matches your outcome and design, and understand what belongs in a different methods category.

Use this section when your question is how an outcome changes with one or more predictors and you need a model that estimates that relationship.

This section is for outcome-predictor modeling, not for simple group-difference testing, contingency-table association tests, or time-to-event analysis. If your main question is whether groups differ, whether categorical variables are associated, or when an event happens, start with Group Comparison, Categorical and Association, or Survival Analysis instead.

Regression methods are useful when you need to estimate slopes, odds, non-linear dose-response patterns, or clustered and repeated-measures effects while accounting for covariates. Licklider helps you choose the model family that matches your outcome type and design, then directs you to the leaf page for assumptions, interpretation, and reporting.

Child pages

Start here

  1. Start with Linear Regression (OLS) when your outcome is continuous and you want to estimate a linear relationship with one or more predictors.
  2. Start with Correlation Analysis when you need Pearson and Spearman correlations, primary selection from residual normality, and how they appear on scatter or regression figures.
  3. Start with Logistic Regression and AUC/ROC when your outcome is binary and you need probabilities, odds ratios, discrimination, or classification performance.
  4. Start with Non-linear Regression and IC50/4PL when your response follows a curved dose-response or saturation pattern that a straight line cannot capture.
  5. Start with GLMM: Gaussian and Binomial when observations are clustered, grouped, or partially dependent and you need fixed and random effects in the same model.
  6. Start with Repeated Measures and Mixed Models when the same subject contributes multiple measurements across time, condition, or replicate structure.
  • Use related links below when your modeling question also needs supporting guidance from setup, checks, figures, or reporting.