Motivation | A worked example | Screening with FDR control | Block decomposition | Classical CNC panels | Perturbation schemes | A practical workflow | Why the guarantee holds | Semiparametric fits | Smoothing-parameter selection

Motivation | A worked example | Screening with FDR control | Block decomposition | Classical CNC panels | Perturbation schemes | A practical workflow | Why the guarantee holds | Semiparametric fits | Smoothing-parameter selection

Introduction | The data | Fitting the model | Local influence diagnostics | Deletion analysis | Model adequacy | Density shapes | Reproducing the full study | References

Overview | The data | Fitting a simplex regression model | Influence diagnostics | The bootstrap goodness-of-fit test | Half-normal plot with simulated envelope | Convenience plot methods | Next steps

The ammonia application | One-call reproduction | Parameter estimates (Table 5 of the paper) | Goodness-of-fit results (Table 6 of the paper) | Step by step | References

The PBSC application | The data | One-call reproduction | Parameter estimates | Goodness-of-fit results | Step by step | References

Why log-cumulants? | A first analysis | Comparing candidate models | Where to go next

The three nested statistics | Why the bootstrap? | Full model comparison

Empirical size | Empirical power | Reproducing the figures

Log-cumulant diagram | Kurtosis-skewness diagram | Coefficient-of-variation diagram | All three at once | Several datasets on one diagram

Introduction | Installation | Independent observations — glm | Data | Fitting the model | Delta method — conditional standardisation | Delta method — marginal standardisation | Custom reference values | Forest plot | Bootstrap confidence intervals | Clustered / multilevel data — glmer (lme4) | Longitudinal data — GEE via geepack | Complex survey data — svyglm (survey) | Comparing OR and PR | Methodological notes | Conditional standardisation | Marginal standardisation | Delta method | Session information | References

Example 1 — Low Birth Weight (LBW): clustered binary data | Data | Independent GLM (ignoring clustering) | GEE — accounting for within-mother correlation | Mixed model (random intercept per mother) | Example 2 — Thailand Education Study: multilevel data | Independent GLM | Mixed model | Example 3 — Toenail Infection Trial: longitudinal data | GEE | Example 4 — UIS Drug Treatment Study | GLM — independent observations | OR vs PR comparison | Bootstrap CIs | Example 5 — Downer Cow Survival | GLM | Example 6 — Titanic Survival | GLM — OR vs PR | Forest plot | Key comparison for Titanic | Summary | Session information