HOME › TEACHING › EXPERIMENTAL DESIGN

Experimental Design in Biology.

BIOL 683. A graduate seminar in experimental design that prepares PhD students to be rigorous, fast, and AI-savvy scientists. R programming through "vibe coding", a carefully sequenced dataset progression, and statistical methods from t-tests through Bayesian MCMC.

Course information

Number: BIOL 683
Title: Experimental Design in Biology
Semester: Fall 2026
Time: TR 8:00 to 9:15 AM
Location: ILCB 237
Credits: 3
Syllabus: Official TAMU syllabus

Instructor

Name: Heath Blackmon
Office: BSBW 425
Email: blackmon@tamu.edu
Hours: TR 9:30 to 10:30 AM, or by appointment

Goals and philosophy

This course prepares PhD students to be rigorous, fast, and AI-savvy scientists. Students learn R programming through a "vibe coding" approach, building intuition and working alongside AI tools, while progressing through a carefully sequenced series of datasets that grow in complexity. Each week builds skills in data visualization, hypothesis testing, model building, and critical evaluation of both statistical methods and AI-generated output.

By the end of the course, students understand when and how to apply t-tests, chi-square tests, linear models, generalized linear models, mixed effects models, dimensional reduction techniques, and Bayesian MCMC approaches, not as isolated methods but as part of a cohesive analytical toolkit.

Learning objectives

Develop practical proficiency in R for data visualization, manipulation, and analysis.
Learn "vibe coding": using AI as a coding partner while understanding the logic behind your analyses.
Progress through increasing complexity in experimental design and statistical methods.
Master hypothesis testing and model selection frameworks.
Apply mixed effects models, dimensional reduction, and Bayesian methods to real biological data.
Critically evaluate statistical output, model assumptions, and AI-generated suggestions.
Communicate results clearly through code, visualizations, and scientific writing.
Practice responsible and ethical use of AI in research.

Resources and tools

Guides and references

Prompting guide Best practices for working effectively with AI tools in data analysis and coding.

Statistical methods guide Reference for when to use different statistical approaches.

Statistical methods: deep dive Detailed explanations and mathematical foundations of key methods.

Method choice tool Interactive guide to help you choose the right statistical test for your data.

Datasets

Course datasets Curated biological datasets used throughout the course, organized by topic and complexity.

Interactive tools

Distribution simulator Interactive R app to explore probability distributions and sampling behavior.

T-test simulator Interactive app to visualize how t-tests work with different sample sizes and effect sizes.

Course schedule, Fall 2026

A dataset-focused progression from foundational R and vibe coding through advanced statistical methods.

Wk	Dates	Topic	Dataset focus
1	Aug 24 to 30	Orientation and the AI scientist	Course philosophy, vibe coding intro, AI tools setup
2	Aug 31 to Sep 6	Foundations of R and vibe coding	Base R syntax, reading data, using AI as coding partner
3	Sep 7 to 13	Data visualization with base R	Morphometric measurements: plotting, histograms, boxplots
4	Sep 14 to 20	Statistical thinking and distributions	Ecological count data: probability, distributions, hypothesis framework
5	Sep 21 to 27	T-tests and permutation tests	Field experiment data: comparing two treatment groups
6	Sep 28 to Oct 4	Test week 1: AI-critical thinking exam	Synthesizing weeks 1 to 5; evaluating AI-generated analyses
7	Oct 5 to 11	Chi-square and tests of proportions	Categorical data: genetic cross ratios or species surveys
8	Oct 12 to 18	Linear models part 1: ANOVA and regression	Multi-factor experimental data: comparing multiple groups and continuous predictors
9	Oct 19 to 25	Generalized linear models and model selection	Count/binary response data: GLMs, AIC-based model selection
10	Oct 26 to Nov 1	Mixed effects models	Longitudinal or multi-site data: random effects, nested designs
11	Nov 2 to 8	Dimensional reduction	Multivariate data: PCA, ordination techniques
12	Nov 9 to 15	Bayesian methods and MCMC	Phylogenetic/evolutionary parameters: priors, posteriors, convergence diagnostics
13	Nov 16 to 22	Choosing your approach: final data discussion	Integrating methods; planning final project analyses
14	Nov 23 to 29	Advanced plotting and work on final	Publication-quality visualizations; no class Thursday (Thanksgiving)
15	Nov 30 to Dec 6	Work on final project	Completion and presentation prep

How the course works

Vibe coding philosophy

"Vibe coding" is about building intuition and confidence in your data analysis skills. Rather than memorizing syntax, you learn to understand the logic of statistical approaches and use AI tools (like ChatGPT or Copilot) as your coding partner, asking them for help with implementation while you focus on the conceptual work. You learn to critique AI suggestions, catch errors, and understand why certain approaches work better than others.

Dataset progression

Each week focuses on a specific biological dataset that demonstrates the week's statistical concept. The datasets increase in complexity: starting with simple morphometric measurements for visualization, progressing through experimental comparisons, categorical data, multi-factor designs, and finally to advanced methods like mixed models and Bayesian approaches. By working with real (or realistic) biological data, students see how methods apply in practice.

Assessment

Test week 1 (week 6): an AI-critical thinking exam where students analyze a dataset and an AI-generated analysis, identifying strengths, weaknesses, and what they would do differently.

Final project: a comprehensive data analysis project where students choose a dataset of interest, apply appropriate statistical methods from the course, and present findings clearly with publication-quality visualizations.