Skip to content

    Statistics for Ecologists: A Frequentist and Bayesian Treatment of Modern Regression Models

    Reviewed by Zhuanzhuan Ma, Assistant Professor of Statistics, The University of Texas Rio Grande Valley on 10/20/25

    Comprehensiveness rating: 5

    The book spans linear models, bootstrap, multiple regression, non-linear modeling (splines/GAMs), GLMs for counts and binary outcomes, zero-inflation, mixed models/GLMMs, GEE, MLE, Bayesian inference, and MCMC/JAGS, with worked R examples and an appendix on reproducible reports. Navigation is strong via a detailed table of contents and chapter structure; key terms are defined in context.

    Content Accuracy rating: 5

    Explanations and formulas are careful and standard, code examples are consistent with current R practice, and Bayesian sections align with mainstream workflows (JAGS, priors, posterior checks). The text distinguishes assumptions and diagnostic checks clearly and avoids overclaiming.

    Relevance/Longevity rating: 5

    Topics map tightly to modern ecological data problems (non-Normal responses, correlation, overdispersion, zero inflation). The frequentist–Bayesian pairing and emphasis on modeling strategy keep the text durable.

    Clarity rating: 5

    The prose is from simple to complex models. Jargon is introduced with context, and code-first illustrations (plus effect plots and residual diagnostics) make abstract ideas concrete.

    Consistency rating: 5

    Terminology and notation are stable across chapters; the modeling framework (design matrices, link functions, hierarchical structures) is applied consistently, which makes cross-chapter learning smooth.

    Modularity rating: 5

    Chapters and subsections can be assigned independently (e.g., GLMs, zero-inflation, mixed models). Exercises and data examples are self-contained, enabling adoption in varied course sequences.

    Organization/Structure/Flow rating: 5

    Flow is logical: Normal-response models --> variable selection/causal thinking --> distributional foundations -->MLE & Bayesian --> GLMs/zero-inflation --> correlated data (LMMs/GLMMs/GEE).

    Interface rating: 5

    The open-access format and permissive license are a plus for teaching adoption.

    Grammatical Errors rating: 5

    Writing is polished and free of grammatical errors; examples and captions are clear and concise.

    Cultural Relevance rating: 5

    Examples are ecologically focused and non-stigmatizing. The tone is professional and inclusive, suitable for diverse classrooms.

    Back