Introduction & prerequisites#
This part builds toward modern statistical modelling and inference workflows used in quantitative biology.
The soil microbiome practical applies the data wrangling and visualisation principles from Data Management and Visualization in R to community data. It can be completed as a group activity before the modelling chapters.
Learning goals#
By the end of this part, you will be able to:
Extend linear models to non-Gaussian responses via generalised linear models (GLMs)
Fit and interpret mixed-effects models for grouped and repeated observations
Fit and interpret logistic regression (binary) and Poisson regression (count data) models
Recognise and handle temporal autocorrelation in time-series data
Fit nonlinear models using nonlinear least squares (NLLS) with appropriate starting values
Construct likelihood functions and estimate parameters via maximum likelihood estimation (MLE)
Combine priors and likelihoods to derive and interpret Bayesian posterior distributions
Compare models using information criteria (AIC, BIC) and likelihood ratio tests
Apply modern statistical workflows for model fitting, checking, and inference in quantitative biology
Prerequisites#
Essential: Solid foundation in linear regression and model diagnostics from Basic Data Analyses and Statistics
Essential: Comfort with R programming, model formulas, and interpreting model output
Strongly recommended: Complete the basic statistics part first — these chapters build directly on regression concepts
Helpful refresher: Pre-work exercises covering probability distributions, likelihood concepts, and numerical optimisation
Recommended pre-work#
If you haven’t done it recently: revisit regression + diagnostics in Basic Data Analyses and Statistics.
If likelihood/probability feels unfamiliar: do the Pre-work exercises before starting.