Design of Experiments Course

Lesson 1 • Core Principles of Experimentation

Covers randomization, replication, and blocking as the three pillars of valid experiments. Links each principle to bias reduction and result reliability.

Lesson 2 • Overview of the DOE Workflow

Maps the end-to-end process from problem definition through analysis and implementation. Gives students a roadmap for all subsequent chapters.

Lesson 3 • Defining Factors, Responses, and Noise

Distinguishes controllable factors, uncontrollable noise variables, and measurable responses. Prepares students to map real problems onto experimental structures.

Lesson 4 • What Is Design of Experiments

Defines DOE, its historical origins, and its role in scientific and industrial problem-solving. Establishes vocabulary used throughout the course.

Lesson 1 • Regression Analysis for Experiments

Applies simple and multiple linear regression to model factor-response relationships. Prepares students for response surface and prediction modeling later.

Lesson 2 • Sample Size and Power Calculations

Teaches how to determine the number of runs needed to detect effects of practical importance. Prevents underpowered experiments and wasted resources.

Lesson 3 • Hypothesis Testing Essentials

Covers Type I and Type II errors, p-values, and confidence intervals for experimental comparisons. Connects statistical decisions to experimental conclusions.

Lesson 4 • Probability and Distributions Review

Reviews normal, t, F, and chi-square distributions as they apply to experimental data. Grounds subsequent hypothesis testing in distributional assumptions.

Lesson 1 • Variance Tests and Equality Checks

Introduces F-test and Levene's test for comparing variances before pooling. Ensures correct test selection based on variance homogeneity.

Lesson 2 • Practical Experiment Execution

Guides students through run-order randomization, data recording, and preliminary analysis. Bridges statistical theory to hands-on experimental practice.

Lesson 3 • Nonparametric Alternatives

Presents Mann-Whitney and Wilcoxon tests for non-normal or ordinal data. Expands the analyst's toolkit beyond parametric assumptions.

Lesson 4 • Multiple Comparison Procedures

Addresses family-wise error rate inflation when testing multiple means simultaneously. Introduces Tukey, Bonferroni, and Dunnett corrections.

Lesson 1 • Residual Analysis and Model Validation

Applies residual diagnostics to verify factorial model assumptions and identify influential points. Ensures model validity before drawing experimental conclusions.

Lesson 2 • Main Effects and Interaction Effects

Defines and calculates main effects and two-factor interactions from factorial data. Teaches interpretation of interaction plots for practical decision-making.

Lesson 3 • Two-Level Full Factorial Concepts

Introduces 2^k designs, effect coding, and the design matrix structure. Establishes the geometric and algebraic basis for factorial experimentation.

Lesson 4 • ANOVA for Factorial Designs

Extends ANOVA to multi-factor experiments, partitioning variance among main effects and interactions. Connects F-statistics to factor significance decisions.

Lesson 1 • Alias Structure and Confounding

Teaches how effects become aliased in fractional designs and how to identify alias chains. Enables informed decisions about which effects can be estimated independently.

Lesson 2 • Analyzing and Interpreting Screening Results

Applies half-normal plots, Lenth's method, and effect sparsity to identify active factors. Guides the transition from screening to optimization experiments.

Lesson 3 • Half-Fraction and Quarter-Fraction Designs

Demonstrates construction and analysis of 2^(k-1) and 2^(k-2) designs for practical screening. Covers fold-over strategies to break aliasing when needed.

Lesson 4 • Plackett-Burman and Supersaturated Designs

Introduces Plackett-Burman designs for screening up to N-1 factors in N runs. Compares their alias structure to regular fractional factorials.

Lesson 1 • Split-Plot Design Structure

Explains whole-plot and subplot factors arising from hard-to-change experimental conditions. Identifies the two error strata and their impact on analysis.

Lesson 2 • Randomized Complete Block Design

Introduces RCBD to remove known nuisance variation from experimental error. Demonstrates how blocking increases precision without additional factor runs.

Lesson 3 • Latin Square and Related Designs

Uses Latin square designs to control two nuisance factors simultaneously with minimal runs. Extends to Graeco-Latin squares for three nuisance factors.

Lesson 4 • Analyzing Split-Plot Experiments

Applies mixed-model ANOVA to correctly test whole-plot and subplot effects. Avoids the common error of using a single pooled error term.

Lesson 1 • Introduction to Response Surface Methods

Defines the response surface concept and the sequential strategy for moving toward an optimum. Positions RSM as the natural follow-up to screening experiments.

Lesson 2 • Path of Steepest Ascent

Teaches the first-order model-based method for moving efficiently toward the optimum region. Covers step-size selection and stopping criteria.

Lesson 3 • Second-Order Model Fitting and Analysis

Fits full quadratic models, interprets coefficients, and validates model adequacy. Prepares students to characterize surface shape and locate stationary points.

Lesson 4 • Central Composite and Box-Behnken Designs

Introduces CCD and BBD as the primary second-order designs for RSM. Compares their rotatability, run counts, and practical suitability.

Lesson 1 • Signal-to-Noise Ratios

Defines and calculates Taguchi signal-to-noise ratios for the three quality objectives. Connects SNR analysis to factor level selection for robustness.

Lesson 2 • Combined Array and Modern Robust Design

Presents the combined array approach that integrates control and noise factors in one design. Compares it to Taguchi's crossed-array approach for efficiency and information.

Lesson 3 • Philosophy of Robust Design

Introduces Taguchi's quality loss function and the goal of minimizing variation around a target. Contrasts robust design with traditional tolerance-based quality control.

Lesson 4 • Robust Design Case Studies

Applies robust design methods to manufacturing, product development, and service process examples. Reinforces the full workflow from noise identification to optimal robust settings.