Foundations of Statistical Decision Making

Develop intermediate statistical decision making for manufacturing professionals with hypothesis testing, t and z tests, anova, and design of experiments (full and fractional factorial, Taguchi methods) through practical case studies.

Examine how to perform z tests, t tests, and ANOVA using Excel and add-ins. Compare software options like Minitab, JMP, and SAS, focusing on design of experiments and interpretation.

Apply basic statistical tools like hypothesis testing and design of experiments to make data-informed decisions, ensure data integrity, and reach clear conclusions.

Download the free pdf slide deck included with this course, featuring all slides discussed, designed with white space for notes to boost learning retention and course value.

Explore how sample-based evidence informs decisions on process changes, supplier selection, safety performance, and defects across outlets, work cells, and six controls, including rival practices.

Apply hypothesis testing and designed experiments to determine if changes in materials, methods, labor, or environment produce significant effects, and compare performance across machines, methods, operators, or suppliers.

Explore how sample averages follow a normal distribution by the central limit theorem and how control charts reveal shifts or instability.

Explore hypothesis testing with the null hypothesis mu = 1000 and the alternative, and learn how rejecting or accepting the null leads to type i and type ii errors.

Compare means using t tests, z tests, and analysis of variance, and explore the F distribution, full factorials, and Taguchi fractional factorial designs.

Use a z test to detect a two-tailed process shift when the population standard deviation is known, comparing a sample mean to the historic 800 units per hour.

Apply t testing with unknown population variation to decide hypotheses, illustrated by rubber hardness and BMI examples using one- and two-tailed tests.

learn how hypothesis testing and experimental design inform statistical decisions, and use Excel, Minitab, and key macros to perform t-tests and z-tests with the Analysis Toolpak.

Perform a t-test in Excel to determine whether a drug alters hemoglobin, using descriptive statistics, the t statistic, and a 0.05 alpha to reject the null hypothesis.

Learn how to perform a paired t test in Excel to compare normal versus low carb diets, using a 0.01 alpha and a two-tailed test.

Apply a paired t-test in Excel to compare cycle diet and control blood pressure. Demonstrate a one-tailed test showing significant blood pressure reduction, leading to rejection of the null.

Explore two-sample t tests on means using real examples with nine students per group, comparing class writing scores and welding scores, and discuss one-tailed versus two-tailed decisions.

Solve five t-test problems applying hypothesis testing, type I and II errors, and the t statistic in Excel across cases like tire design, anxiety, stress, sleep, and boilers.

Use a z test with known sigma to test mu=15; sample mean 14.8 from 50 lines at 1% alpha yields z = -2.83, leading to rejecting the null.

Perform a two-tailed z-test comparing the new nylon coating tooling to the 0.0064 mm standard, using 25 samples; since z = -1.875 is within ±1.96, we accept the null.

Using Excel, perform a one-tailed z test to compare the sample mean of 536 against the population mean 500 (sigma 100, n=25) at 5% alpha, concluding a positive coaching effect.

Demonstrates a one-tailed z test using a sample mean of 212.79, population mean 210, population standard deviation 8.5, and n=42, yielding z=2.13 and rejection of the null.

Demonstrates z tests for known population standard deviation and mean, using two-tailed and one-tailed analyses with alpha 0.05 to accept or reject the null hypothesis on sample means.

Practice solving five z test homework problems with provided solutions, covering college admissions, nitrate effects on mice growth, river alkalinity, height differences between regions, and machine performance.

Learn to test proportions for attribute data by formulating H0 and Ha, using a one-tailed z test at 5% alpha, and calculating sample size, illustrated by ibuprofen side effects.

Calculate the z statistic from ibuprofen side effects data, compare to the 1.65 one-tailed threshold, reject the 3% or less null hypothesis, and justify a warning label.

Review statistical variation and analysis of variance, and evaluate results against tabulated references. Define null and alternate hypotheses, explain type I/II errors, apply t and z tests for means, proportions.

Access a downloadable glossary of terminology that consolidates key terms from the course. Responding to student feedback, it adds terms like logistic regression and Poisson distribution, with a reference pdf.

Learn how t tests, z tests, tests for proportions, and full or fractional factorial experiments with ANOVA reveal how inputs drive outputs and why one-at-a-time studies are inefficient on processes.

Identify experimental factors and factor levels, and analyze response variables, main effects, and interactions, while accounting for experimental error, replication, and independent and dependent variables.

Apply analysis of variance (ANOVA) and the F distribution to test if four thermometers read differently, under a null hypothesis of equal means, using Excel's data analysis tool.

The one-way ANOVA compares four groups—education, business, behavioral science, and Fine Arts—with 32 students’ percent scores on a US history test, and finds no significant difference at alpha 0.05.

Explore one-way ANOVA across four environmental conditions in rats and three drug groups in humans, comparing learning and depression outcomes using F tests.

Explore three homework problems on learning task performance in wild and lab-reared monkeys, and a study linking TV exposure to happiness.

Explore two-way analysis of variance with applied examples: protein breakfast and gender on adolescent performance, and music type with Alzheimer's stage on agitation, using Excel and Key Macro results.

Apply a two-way ANOVA to test whether life satisfaction differs by gender across three age groups, at alpha 0.05, using the provided data and an Excel solution tab.

Explore full factorial and fractional factorial experiments, including the Taguchi method, to identify key factors and interactions among temperature, pressure, and concentration.

Explore the Taguchi fractional factorial method, using orthogonal arrays and linear graphs to test multiple two-level factors with as few as four or eight trials, via L4 and L8 arrays.

We illustrate the Taguchi experiment method with an eight-step example on tube cut length control in air conditioning, using L8 orthogonal array, contrast analysis, and ANOVA to identify key factors.

Taguchi L4 example shows drill speed as the significant factor in drilling quality. Drill type and lubrication do not matter, so use the cheapest drill and save lubricant.

Taguchi L8 experiments examine valve train noise in a luxury car engine, screening six factors and highlighting seat concentricity as a key driver.

Assess a Taguchi designed, fractional factorial study of a laptop fan to reduce noise by evaluating five factors, including fan blade surface and air gap, at two levels.

Apply t-tests, z-tests, and tests for proportions to improve decision making, and use ANOVA with full and fractional factorials to evolve processes.

Conclude the foundations of statistical decision making course by applying intermediate tools to break down data and make accurate manufacturing decisions, with lifetime access and ideas you can apply.

A bonus lecture promoting related Udemy courses in statistical decision making, statistical process control, process capability analysis, supply chain analytics, and manufacturing leadership.