Accelerate your yield improvement and master the data-driven methods that separate hypothesis from guess. Learn to design fast, focused screening experiments that pinpoint the process factors driving yield loss, then translate those findings into experiment plans that your team can execute with confidence.
In three focused modules, you'll design 2-level fractional factorial screening experiments to investigate fab yield excursions (like contact-etch defect spikes or gate-leakage drift), analyze main-effect plots and ANOVA results to rank process factors by their impact, and apply the 5-Why method to convert yield data into testable hypotheses ready for the next experiment cycle. You'll work through scenarios drawn from semiconductor fab reality: contact-etch yield loss, gate-leakage drift, and cross-functional yield reviews where your experiment designs and hypotheses face scrutiny. By course end, you'll be equipped to scope screening experiments that deliver results in weeks, not months, and build defensible hypotheses that point the next experiment in the right direction. Ideal for process engineers, yield engineers, and fab floor leaders tasked with root-cause investigation and continuous improvement in semiconductor manufacturing. By the end of this course, you will be able to: - Design a 2-level fractional factorial screening experiment—including factor selection, run randomization, and response definition—to investigate a specified fab yield excursion (Apply) - Analyze main-effect and interaction plots together with basic ANOVA outputs to rank the top three process factors influencing the experimental response (Analyze) - Apply the 5-Why technique to translate an observed parametric shift into a concise, testable hypothesis statement used to scope an experiment plan (Apply) Working knowledge of yield metrics (defect density, yield loss) and familiarity with fab process steps (lithography, etch, deposition, diffusion). Exposure to statistical fundamentals (mean, variance, normal distribution) is helpful. Access to JMP or Minitab is recommended for hands-on practice with screencasts. No prior DoE experience required.













