Applied Statistics for Engineers

Seminar Overview

Engineers are drowning in data and starving for statistics. This two-day seminar is built from the **NIST/SEMATECH e-Handbook of Statistical Methods** — the multi-section engineering statistics reference maintained by the National Institute of Standards and Technology and SEMATECH — supplemented with the reliability life-data corpus (Weibull and life-distribution methods) for the failure-data questions every quality and test engineer faces. The e-Handbook is the primary source: its exploratory data analysis chapter, probability-distribution gallery, estimation and hypothesis-testing chapters, design-of-experiments and process-modeling sections, and its control-chart and capability material are the backbone of every module.

This is not a mathematics course. It is a decision course: which plot to make first, which test to run, what the p-value does and does not license you to claim, how wide your confidence interval really is, whether your process is capable of the print, and how to design an experiment that answers a question in one round instead of four. Day one covers exploratory data analysis, distributions, estimation, and hypothesis testing; day two covers design of experiments, capability, SPC, and regression — each with worked engineering datasets drawn from manufacturing, dimensional measurement, and reliability practice.

The differentiator: every module is taught from PRIMARY SOURCE material — the NIST/SEMATECH e-Handbook's own methods, worked examples, and case data, plus reliability references for failure-time data — not vendor slide decks. You leave able to run a defensible statistical analysis on your own data and to challenge the statistics you are handed by others.

Ideal Learner

• Quality engineers, process engineers, and manufacturing engineers who own SPC, capability, and validation data
• Design, test, and validation engineers who must design experiments and interpret test results
• Supplier-quality and PPAP engineers reviewing Cpk studies, Gage R&R, and capability submissions
• Reliability and field engineers who analyze failure and warranty data statistically
• Industry segments: automotive OEM/Tier 1, medical devices, electronics, industrial equipment, injection molding and plastics processing

Learning Objectives

• Perform structured exploratory data analysis — histograms, box plots, run charts, lag plots, and normal probability plots — and state what each is diagnostic of before running any test
• Select the correct probability distribution (normal, lognormal, exponential, Weibull, binomial, Poisson) for an engineering dataset and justify the selection with a probability plot
• Construct and interpret confidence, prediction, and tolerance intervals for means, variances, and proportions — and explain the difference between the three on the shop floor
• Formulate and run the correct hypothesis test (t-test, F-test, chi-square, ANOVA), set significance and power, and size the sample before the test, not after
• Plan a full or fractional factorial designed experiment, read main effects and interactions, and state the experiment's answer in engineering units
• Compute and interpret Cp, Cpk, Pp, and Ppk against one- and two-sided specifications, including non-normal capability via Weibull analysis
• Select and build the correct control chart (X̄-R, individuals, p, np, c, u, EWMA) and distinguish common-cause from special-cause signals
• Build a least-squares regression model with honest residual analysis and defend its limits

Consulting Sessions

Seminar attendees can sign up for individual consulting sessions with the instructor. Sessions are free for registered attendees, first-come first-served — sign up when registering by calling 248-539-0473 or during the seminar.