Experimental Design Course

Lesson 1 • Core Experimental Terminology

Introduces variables, units, and treatments as precise technical terms. Accurate vocabulary prevents misinterpretation throughout the course.

Lesson 2 • The Role of Hypotheses

Covers null and alternative hypotheses and their logical structure. Connects hypothesis formation to design choices made later in the course.

Lesson 3 • Validity and Reliability Basics

Distinguishes internal from external validity and defines measurement reliability. These concepts govern every design decision covered in later chapters.

Lesson 4 • What Is an Experiment

Defines experiments versus observational studies and surveys. Establishes the chapter's conceptual baseline for all subsequent design decisions.

Lesson 1 • Classifying Variables

Categorises continuous, discrete, nominal, and ordinal variables. Correct classification determines which statistical methods apply in later chapters.

Lesson 2 • Blinding and Placebo Techniques

Explains single-blind, double-blind, and placebo controls for reducing bias. These techniques directly address observer and participant expectancy effects.

Lesson 3 • Control Strategies

Presents holding constant, matching, and statistical control as complementary techniques. Each strategy reduces noise and strengthens causal inference.

Lesson 4 • Identifying Confounding Variables

Defines confounds and explains how they bias treatment estimates. Recognizing confounds is prerequisite to the randomisation strategies in Chapter 3.

Lesson 1 • Adaptive Randomisation

Presents response-adaptive and covariate-adaptive allocation schemes. These advanced methods optimise balance or outcomes during ongoing trials.

Lesson 2 • Documenting and Auditing Randomisation

Establishes procedures for recording, verifying, and reporting random assignment. Transparent documentation supports reproducibility and regulatory review.

Lesson 3 • Simple and Systematic Randomisation

Covers coin-flip, random-number, and systematic interval methods. Students apply these foundational techniques before advancing to restricted forms.

Lesson 4 • Restricted Randomisation Methods

Introduces block, stratified, and cluster randomisation to improve balance. Each method addresses specific imbalance risks identified in earlier sections.

Lesson 1 • Sampling Strategies Overview

Compares probability and non-probability sampling and their inferential implications. Sampling strategy selection directly affects external validity discussed in Chapter 1.

Lesson 2 • Sample Size Calculation Methods

Demonstrates formula-based and simulation-based sample size estimation. Students apply these methods to means, proportions, and regression contexts.

Lesson 3 • Statistical Power Fundamentals

Defines power, Type I error, and Type II error and their interrelationships. Understanding power is prerequisite to all sample size calculations in this chapter.

Lesson 4 • Pilot Studies and Feasibility Testing

Uses small-scale pilots to refine effect size estimates and logistics. Pilot findings feed directly into revised sample size calculations before full deployment.

Lesson 1 • Crossover Designs

Assigns each subject to multiple treatments in sequence to reduce variability. Carryover effects and washout periods are central design considerations.

Lesson 2 • Completely Randomised Design

Covers single-factor CRD structure, assumptions, and analysis linkage. This simplest design serves as the reference point for all more complex structures.

Lesson 3 • Randomised Complete Block Design

Introduces blocking to remove known nuisance variation from error. RCBD extends CRD by adding the blocking concept introduced in Chapter 3.

Lesson 4 • Latin Square Design

Controls two nuisance factors simultaneously using a square arrangement. Students apply Latin squares when two blocking dimensions are identifiable.

Lesson 1 • Fractional Factorial Designs

Reduces run count by confounding high-order interactions with main effects. Students construct and interpret resolution III, IV, and V fractions.

Lesson 2 • Two-Factor Factorial Design

Introduces the 2×2 and general two-factor factorial structure and interaction concept. Interaction interpretation is the central skill built throughout this chapter.

Lesson 3 • Higher-Order Factorial Designs

Extends factorial logic to three or more factors and higher-order interactions. Students manage the complexity of three-way and four-way factorial structures.

Lesson 4 • Two-Level Full Factorial Designs

Focuses on 2^k designs for screening and optimisation contexts. Coded factor levels and effect estimation are core computational skills.

Lesson 1 • Fitting and Validating Response Surface Models

Fits second-order polynomial models and assesses their adequacy. Model validation ensures predictions are reliable before optimisation decisions are made.

Lesson 2 • Moving from Screening to Optimisation

Explains the sequential experimentation strategy from screening to fine-tuning. This section bridges factorial designs from Chapter 6 to response surface methods.

Lesson 3 • Central Composite Design

Constructs CCD by augmenting a 2^k factorial with axial and center points. CCD is the most widely used second-order design for optimisation.

Lesson 4 • Locating and Interpreting Optima

Uses canonical analysis and contour plots to find stationary points. Students distinguish maxima, minima, and saddle points in the response surface.

Lesson 1 • Matching Analysis to Design

Aligns statistical models with the design structure used to collect data. Mismatched analysis invalidates conclusions regardless of design quality.

Lesson 2 • Transparent Reporting Standards

Applies structured reporting guidelines to document design and analysis decisions. Transparent reporting enables peer review and independent replication.

Lesson 3 • Multiple Comparisons and Error Control

Applies correction procedures to control familywise and false discovery rates. Students select appropriate corrections based on the number and type of comparisons.

Lesson 4 • Reproducibility and Open Science Practices

Embeds data sharing, code documentation, and open materials into the workflow. These practices address the replication crisis and build long-term credibility.