Design-Expert v25.0
- What’s New
- Installation
- Getting Started
- Tutorials
- Designs
- Pre-Experiment
- Analysis
- Optimization
- Post Analysis
- Python Integration (Stat-Ease 360® only)
- Gaussian Process Models (Stat-Ease 360® only)
- Advanced Topics
- Logistic Regression
- Candidate Point Creation
- Complex Constraints
- Convert Coded response surface model to Actual
- Converting Mixture Models from Pseudo to Real
- Convert a quadratic L_Pseudo mixture model to Real
- Convert a quadratic U_Pseudo mixture model to Real
- Covariates and Additional Explanatory Variables
- Design Augmentation
- Equation Entry
- Exponential Notation
- Extrapolating a Mixture Design
- Extrapolating a Response Surface Design
- Foldover
- Fraction of Design Space Computations
- Fraction of Paired Design Space computations
- Interpreting Models
- Mean Bias Correction
- Multiple Linear Constraints
- Optimality Criteria
- Optimal Exchange Methods
- Propagation of Error
- Standardized and Normalized Factorial Effects
- Start the Design From Existing Data
- Restricted Maximum Likelihood (REML) vs Maximum Likelihood (ML) Analysis
- Hints and FAQs
- Technical Support
- Statistical Support
- Credits
Design-Expert
» Advanced Topics
» Logistic Regression
» Fit Statistics
» Pseudo R-Squared
» McFadden Pseudo R-Squared
McFadden Pseudo R-Squared
The McFadden pseudo \(R^2\) compares the log likelihood of the full model (\(L\)) to the log likelihood of the model with just the intercept (\(L_0\), the null model).
\[R^2 = 1.0 - \frac{\ln(L)}{\ln(L_0)}.\]
References