Quality Engineering Statistics (2026)

Explore the different data types used in quality engineering statistics, including variable and attribute data, continuous and discrete data, and qualitative versus quantitative attributes.

Explore the four measurement scales—nominal, ordinal, interval, and ratio—and how they define data, ranking, equal spacing, and the significance of zero, including absolute zero.

Simplify data collection with data coding. It reduces writing and digits while preserving meaning, for manual and automated collection, with examples like subtracting 10,000 or multiplying by 1000.

Preserve data integrity by refusing to falsify or eliminate data, recording context and environment, avoiding excessive rounding, screening mis-recordings, and securely storing raw data for future analysis.

Explore data visualization techniques that transform raw data into charts like pie charts and box-and-whisker plots, revealing patterns and enabling clear communication to technical and non-technical audiences.

Explore a stem and leaf diagram that visualizes test scores in Excel, with a 30-item data set sorted largest to smallest and stems like 90, 80, 70.

Learn to build histograms in Excel, understand automatic and custom binning, and generate a frequency distribution using the data analysis tool pack for clear data visualizations.

Dashboards consolidate data visualizations to monitor quality metrics and KPIs in real time. Visualizations empower organizations to spot trends, understand performance, and tell the data story to stakeholders.

Explore descriptive statistics to describe past data through central tendency and dispersion, learn the arithmetic mean, mode, variance, and standard deviation, and differentiate sample from population measures.

Explore how histograms reveal dispersion in quality engineering statistics by showing the same mean and sample size but different standard deviations and bin patterns.

Explore measures of dispersion by calculating sample variance and sample standard deviation in Excel, comparing the hand calculation with the shortcut using var.s and stdev.s.

Describe the shape of data using frequency distributions and histograms. Learn ungrouped and grouped distributions, cumulative and percent frequencies, and the use of bins.

Demonstrate categorical frequency distributions for defect types, and apply Pareto analysis with cumulative frequency, cumulative percentage frequency, and an ordered histogram to focus on the top defects.

Explore non parametric statistics by learning median, quartiles, and IQR, and apply them in Excel to analyze central tendency and dispersion.

learn to create a box and whisker plot in excel, and interpret min, max, quartile one, quartile three, median, and the interquartile range to assess dispersion and detect outliers.

Strengthen quality engineering statistics skills with practice exercises on data coding, data visualization, mean, median, mode, and standard deviation, plus Excel datasets and answer keys.

Download a glossary of about 110–115 engineering statistics terms related to quality, including measures of central tendency and distributions, and preview concepts like normal distribution and Weibull.

Explore a downloadable collection of 19 excel worksheets covering quantitative concepts and analytical methods for quality engineering statistics, with practice exercises and tutorials to apply at your workplace.

Compute the population mean mu and variance in Excel by averaging 30 data points to obtain mu = 0.8311 and apply the variance formula, yielding 0.0005.

Describe how variance measures dispersion from the mean, and how the standard deviation is its square root; relate these to the normal distribution, PDF, and CDF.

Explore how the sample mean and sample standard deviation estimate population parameters, using x-bar and s with n and n-1, from a 30-item population and a six-item sample.

Analyze discrete data, such as defects per unit, using descriptive statistics and histograms in Excel. Interpret mean, median, mode, and range to summarize the defects-per-unit distribution.

Analyze continuous data by linking probability to the area under the probability density function, illustrated with a 100-diameter sample, descriptive statistics, bins, and a histogram.

Explore central-tendency measures by comparing mu and x-bar, and describe median, mode, skewness, range, normal distributions, and kurtosis to detect outliers and data tails.

Explore sampling plans—single, double, multiple, and sequential—for attributes or variables, guided by AQL and LTPD, and learn how random sampling and OC curves govern inspections.

Compute the expected value by weighting outcomes with their probabilities and summing xi pi, using casino games to illustrate long-term house advantage.

Enumerative studies count and describe a population at a specific point in time, while analytic studies analyze numerical data and process variables to improve future performance.

Evaluate the validity of statistical conclusions by ensuring reliable measurements, guarding against bias, preventing rounding errors, preserving data order, and maintaining traceable classifications.

Explore threats to statistical conclusion validity by examining type one and type two errors, and learn to reduce them with alpha settings, larger sample sizes, and larger effect sizes.

Examine how restriction of range distorts statistical validity, illustrating pencil whipping, data trimming, and biased correlations in examples with students, tutoring, and athletes, plus Excel regression.

Uncover how violated assumptions threaten statistical validity by testing normality, evaluating skewness and kurtosis, and applying chi-square and anderson-darling goodness-of-fit with observed versus expected counts.

Assess post position effects using a chi-square goodness-of-fit test. For 144 races, a chi-square of 16.33 (df=7) is below 18.47, so we accept the null of no effect.

Apply the Anderson-Darling test to assess if data follow a specified distribution, with tail-focused sensitivity. In a 100-diameter sample, ad=0.440 is below 0.752, accepting normality.

Explore robust statistics that stay reliable despite deviations, outliers, and diverse distributions, guided by the central limit theorem for t tests and ANOVA.

Clarifies hypothesis testing, null and alternative hypotheses, and alpha and beta errors, showing how sample size, variability, and effect size influence statistical power with a practical power calculator example.

Identify how a poorly fit statistical model undermines conclusions by showing symptoms like large residuals, assumption violations, and poor predictive performance, and review key distributions to test model fit.

Explore foundational probability concepts by defining random variables, experiments, and sample spaces, and distinguish discrete versus continuous variables, independence, and basic probability calculations with coins and dice.

Apply the multiplication rule of probability to independent events with and without replacement, using a ten marble bag to compare 3/10 × 3/10 and 3/10 × 2/9.

Use Venn diagrams to compare unions and intersections: A ∪ B includes all points in A or B, A ∩ B includes only their overlap, and A's complement is points not in A.

Apply the addition rule for not mutually exclusive events and distinguish it from mutually exclusive cases. Use P(A∪B)=P(A)+P(B)−P(A∩B) with dice and math and history probabilities to compute joint outcomes.

Explore conditional probability and dependence in quality engineering statistics, showing how the probability of an event given another depends on the reduced sample space when sampling without replacement.

Explore the multiplication rule for independent events using relay A 0.9 and B 0.8, then apply complements and the addition rule to find at least one missile hit with 0.9975.

Explore the sequence tree diagram to compute the probability of at least one hit, comparing independent and dependent trials in a two-missile reliability example.

Explore key probability rules for events A and B, including mutually exclusive events and independence, as presented in a practical table to solve problems.

Practice problems reinforce section b probability concepts, including die rolls and coin flips, with attached documents and an answer key to solidify your foundation.

Define probability distributions, continuous and discrete, and describe outcomes with PDFs and PMFs. Explore normal (Gaussian) distributions with mean and standard deviation, exponential and Weibull distributions, and sampling distributions.

Learn to use norm.dist in excel to find cumulative probabilities for the normal distribution with mean 100 and sd 15, including below 140, below 120, and between 120 and 140.

Explore the Weibel distribution and its beta and eta parameters, their effect on pdf shape and failure rate, and compare discrete distributions like binomial, Poisson, multinomial, and hypergeometric.

Explore the Poisson distribution as the essential discrete model, its approximation and exact form, and relate it to hypergeometric, uniform, multinomial distributions, sampling distributions, and hypothesis testing.

Explore the Poisson distribution in Excel with copper wire defects and taxi arrivals. Learn to compute exact and cumulative probabilities using spreadsheet formulas and unit conversions.

Use the t distribution for estimating means, t tests, and confidence intervals; apply the F distribution with one-way and two-way ANOVA and chi-square tests for independence and goodness of fit.

Section C offers four practice problems on probability distributions to estimate the expected bolt weight, its standard deviation, and the percentage within specified weight ranges.

Explore point estimates and confidence intervals to infer population parameters from samples, focusing on means, variances, and standard deviation, with sample mean x-bar and sample variance s^2.

Explore point estimates in Excel by calculating the sample mean, variance, and standard deviation, and contrast them with population equivalents, highlighting n versus n-1 denominators and sanity checks.

Construct and interpret confidence intervals for the population mean using known and unknown standard deviations, applying z and t factors, sample size, and the confidence level.

Compute confidence intervals for population proportions via the normal approximation using p_S, n and z; illustrate 95% CI for cell phone ownership, and apply chi-square intervals for standard deviation.

Apply confidence interval methods in Excel to estimate the population mean from a 30-point sample, using the sample standard deviation and alpha to compute upper and lower 95% confidence limits.

Describe how standard error of the mean measures variation of sample means from population mean. Show that larger samples reduce error and relate standard deviation to sem and 95% CI.

Explore the standard normal distribution, define the z score and z table, and use them to locate 95% confidence intervals and left-tail probabilities.

Construct a two-sided confidence interval for population variance using the chi-square distribution with n-1 degrees of freedom, illustrated by a 20-sample example and the sample variance s^2.

Compute a two sided 95% confidence interval for variance using chi square coordinates with 19 degrees of freedom. Use alpha/2 = 2.5% and chi square distribution tables to determine interval.

Learn how tolerance intervals differ from confidence intervals and compute a two-sided interval that covers 90 percent of the population with 99 percent confidence.

Understand hypothesis testing as a straightforward method to detect significant changes in materials, methods, labor, or environment. Compare performance across machines, suppliers, or operators to reveal real differences.

Understand how the statistical view of data in statistical process control uses sample averages, the normal distribution, and standard deviation to detect shifts and minimize variation around targeted process streams.

Explore sampling from distributions, tail samples versus the center, and frame hypotheses—null and alternative—using examples like baking buns and inspection, plus type I/II errors and decision rules.

Explore type i and type ii errors in hypothesis testing, defined by alpha and beta, with the operating characteristic curve and percent defective guiding acceptance or rejection of the null.

Compare statistical significance with practical significance, and learn how sample size, variability, and effect size influence real-world importance beyond hypothesis testing with null and alternative hypotheses.

Define clear objectives, control factor effects, and avoid bias to ensure sound experiments; prefer balanced designs for greater power and stable variance when testing means.

Compare means with t and z tests, and analyze variance with anova using the f distribution; explore full and fractional factorial designs, including the Taguchi method.

Demonstrates t-test calculations for hypothesis testing, using sample means, deviations, and critical values to decide null rejection for two-tailed and one-tailed tests, with rubber hardness and BMI examples.

Learn to perform t tests in Excel, compute the mean, standard deviation, standard error of the mean, degrees of freedom, critical t, t value, and p value using Excel formulas.

Apply the z test for means to detect a process shift, with known historical mean and standard deviation, following six-step, two-tailed hypothesis to decision rule.

Apply z-tests to evaluate hypotheses using two-tailed critical values of ±1.96. Determine acceptance or rejection regions and compute z statistic from sample means to decide whether the process changed.

Use two-tailed z-tests to compare sample means with claimed strengths, interpreting alpha, critical values, and z statistics to accept or reject the null hypothesis in manufacturing cases.

Perform a z test in Excel using mu 15, alpha 0.01, n 50, x̄ 14.8, s 0.5; calculate z ≈ −2.83 and reject null when z exceeds the critical −2.58.

Apply z tests of proportions to determine if a process change affects a proportion, using null and alternative hypotheses, one-tailed testing, and a 5% significance level.

Assess statistical conclusion validity by ensuring data accuracy, adequate sampling procedures, appropriate statistical tests, and reliable measurement procedures; minimize emotional bias and data misreporting.

Explore how statistical robustness supports tests under violated assumptions, leveraging the central limit theorem for t procedures and anova, with attention to p-values and sample size.

Explore goodness of fit by testing normality and using chi-square methods to compare observed and expected frequencies, illustrating how to reject or fail to reject the null hypothesis.

Identify threats to statistical conclusion validity by examining restriction of range and data cherry-picking, and illustrate how data selection can distort correlation and regression results.

Examine low statistical power, the null and alternative hypotheses, type I and II errors, and how alpha–beta trade-offs shape power and the operating characteristic curve.

Define statistical power and its relation to type II error, alpha, and effect size. Show calculating required sample size with z-scores for a two-tailed test and 80% power.

Explore how a normal probability plot assesses normality by plotting ordered data against standardized normal values in Excel, using steps from sorting data to evaluating linearity.

In this anova example with four thermometers measuring hydroquinone’s melting point, compute sums of squares, correction factors, and the f distribution critical value to test treatment effects.

Learn to perform a one way ANOVA in Excel using the data analysis toolpak, compare groups with alpha 5%, and interpret F and F critical values.

Apply a two-way analysis of variance to compare thermometers and analysts, using F values and 5% critical values to accept thermometers are the same and reject analysts are the same.

Perform a two-way ANOVA in Excel using the data analysis tool pack to compare four thermometers and three appraisers, interpreting F statistics and critical values.

Explore contingency tables and the chi-square test to assess independence, compare observed versus expected frequencies, and conclude that heating cycle length affects nylon bars' toughness.

Practice problems in section d reinforce statistical decision making with four exercises, an answer key, and a downloadable excel file to apply concepts to real workplace problems.

Explore linear regression and least squares to form a regression equation. Use the predictor equation to estimate tool life from cutting speed and interpret the intercept and slope.

Estimate the error variance, compute the t statistic to test beta one, and compare with the f statistic, then interpret the 95% confidence interval for beta one.

Explore the coefficient of determination (R squared) as the explained variation in linear regression, its link to the correlation coefficient (r), and practical confidence intervals with examples.

Learn a quick graphical approach in Excel to obtain a linear regression equation and R-squared for a cutting speed versus tool life dataset, using scatter plots and a trend line.

Explore how time series data collected over time model dependencies, specify and fit models, diagnose fit, and forecast future values using techniques like moving averages and seasonality analysis.

Work through four problems on the relationship between variables, use the included F distribution table, and access two downloadable resources—a homework problem and an answer key—to reinforce understanding of correlations.

Explore nonparametric methods that do not assume a specific data distribution, compare them to parametric tools, and learn practical alternatives like Spearman correlation and Wilcoxon tests with Excel templates.

Define product and process parameters and explore how statistical robustness guards conclusions under non-ideal data, using random sampling, central limit theorem concepts, and t tests and ANOVA.

Explore statistical inference by comparing parametric and non-parametric methods to decide whether observed sample differences reflect same population, using null and alternate hypotheses and normality, independence, and variance assumptions.

Explore nonparametric statistics with practical Excel worksheets, including sign tests and Wilcoxon signed rank, alongside parametric methods. Follow Mike through tools, case studies, and downloadable resources to build hands-on expertise.

Review parametric tools, including t tests for means, t distribution, binomial probabilities, ANOVA with the F distribution, and Pearson's r, and contrast with non-parametric distribution-free methods.

Use the sign test to compare two related samples when parametric tests can’t be used. It focuses on the direction of changes and median differences, illustrated by bolts quality example.

Explore the Wilcoxon signed rank test, a non-parametric alternative to the t test for paired data, and learn the six-step process to compute the W statistic and test the null.

Apply the Mann-Whitney U test (Wilcoxon rank sum) to compare two independent groups on a continuous variable with a non-parametric, two-tailed 5% significance level; testing for no difference.

Apply the Friedman test, a non-parametric method for more than two related samples, using rankings to assess whether athletic performance changes over time.

Apply the Kruskal-Wallis H test, a non-parametric extension to compare two or more independent samples in one-way classifications, testing if they share a distribution via chi-square with k-1 dof.

Explore Spearman's rank correlation coefficient, a nonparametric measure of monotonic relationships that ranks data to estimate correlation, and compare it with Pearson's parametric method using a ten-athlete example.

Apply statistical process control by exploring seven control charts—three variable and four attribute—building and interpreting them in Excel, and mastering short run spc and pattern interpretation.

Explore statistical process control, control charts, and Taguchi's continuous quality loss function to minimize variation around a targeted process and improve quality.

Master control chart terminology, including x bar, x double bar, and range, and distinguish specification limits from control limits while reviewing run charts and chart basics.

Chart a focused set of characteristics, prioritizing safety, critical-to-quality items, and those tied to customer complaints. Use key process input and output variables to drive quality and process control.

Explore common cause and special cause variation in control charts, and learn rational sub grouping for x-bar charts to minimize within-subgroup variation and maximize between-subgroup variation.

Explore the x-bar and r chart via a saw-cut pipe case study, using 150 samples and five samples per 50 pieces, and plan to build it in Excel.

Develop x-bar and r control charts in Excel from a 75-sample study of 5,500 saw-cut pieces, using a five-sample, 15-interval plan to estimate population limits for quality engineering statistics.

Compare the X-bar and R control charts to show a process in spec and in control, with 99.7% of values within control limits and a narrow, centered range.

Explore x-bar and s charts as dispersion-focused alternatives to r charts, using standard deviation to measure variation and establish control limits with a minimum of nine samples.

Learn to build x bar and s charts in Excel with 12 samples, calculating x double bar and s control limits with a three, b three, b four constants.

Explore the ixmr chart, an individual X and moving range chart for 100% or infrequent production, and learn how moving range uses current and previous values, including mr3 in Excel.

Construct IX and moving range charts for individual readings using absolute differences, X bar, R bar, and constants E two and D four, with MA three or MA five options.

Explore p charts for attribute data, measuring the proportion of good versus no good across lots; learn to set control limits in Excel and compare varying lot sizes.

Learn to build a P chart in Excel from real lot data, calculate proportion defective, average P and N, determine upper and lower control limits, and interpret unusual points.

Explore the np chart for counting defects in a fixed lot of 500 pieces, contrasting it with the p chart and focusing on attributes and whole-number defect counts.

Learn to build an NP chart in Excel by computing NP bar (average defects per lot) and P bar. Then set upper and lower control limits to spot out-of-control points.

Explore the C chart for counting defects per single sample, based on the Poisson distribution, with applications to large components and vehicle paint blemishes.

Builds a c chart for defects with sample size one, computing C-bar and upper and lower control limits via three sigma, handling negative LCL, and distinguishing special from common-cause variation.

Apply the u chart to monitor defects per standard unit when sample size varies, using Poisson-based calculations for bulk processes in quality engineering statistics.

Compute a u chart in Excel to monitor fabric defects per square meter, using standardized unit measures, Poisson-based upper and lower control limits, and graphical interpretation.

Interpret control charts by recognizing patterns and identifying potential root causes, applying the art of quality engineering to attribute and variable charts alike.

Learn to interpret control charts, identify out-of-control signals like jumps and shifts, and recognize recurring patterns and external influences shaping process variation.

Learn to interpret control charts by spotting trends, truncation, and lack of discrimination, and apply Western Electric rules to detect out-of-control conditions and guide investigations.

Explore a practical cheat sheet for control chart selection in quality engineering statistics, distinguishing variable versus attribute data and guiding chart choices with sample size and formulas.

Apply short run spc to make-to-order production with lean and jit; use x-bar and range charts, deviation from nominal, z-chart transformations, and monitor cpk above 1.0 to detect variation.

Wraps up a comprehensive statistical process control section with practice exercises on seven control charts, including individual moving range, U, NP, and fraction defect charts, with downloadable Excel templates.

Explore how process capability analysis compares specification limits to the process output, using product and process variables to express the voice of the process versus the customer.

Specification limits are the voice of the customer, guiding engineers from drawings to process capability analyses, with bilateral and unilateral limits defined by upper and lower specs.

Learn sampling frequency and three methods—100% inspection, random sampling, and interval sampling—to assess process capability and support pap for fit and function features.

Explore three options for process capability analysis—from prototype measurements and run charts to inferential sampling—covering key indices like P, Pk, CP, and CPK with math and Excel.

Examine the normal distribution and sampling, defining the mean, x-bar, and standard deviation while noting that a minimum 30-piece sample provides reliable estimates; smaller samples use the t distribution.

Learn the arithmetic mean, its population mu vs. sample X bar distinction, and how to compute it by hand or in Excel using sum, count, and average.

Perform process capability analysis by using histograms to compare three processes A, B, and C against the 25.00 ± 1 mm specification, showing max, min, and range.

Enable the Excel data analysis add-in (Analysis ToolPak) in Excel 2016 by going to File > Options > Add-ins, then check Analysis ToolPak to access data analysis tools.

Explore how histograms reveal underlying distributions, and learn why the normal distribution forms the bell curve for continuous data, underpinning statistics in quality engineering.

Compute population and sample standard deviation to measure data spread by subtracting mu from each x_i, squaring, summing, dividing by N or n-1, then taking the square root.

Explore normal distribution curve, its mean and sigma, and how 68.3%, 95.4%, and 99.7% of values fall within one, two, and three sigma, defining six sigma process range and Cpk.

Learn to assess whether raw data follow a normal distribution by overlaying the normal curve on histograms. Recognize bimodal or non-normal patterns and exponential distribution possibilities.

Conduct a first capability analysis on a 100-piece batch, using a histogram and descriptive statistics to introduce Pp and Ppk.

Learn how Pp and Ppk evaluate process capability from a population perspective, using mu, sigma, and spec limits to compare the specification range with the process range.

Explore calculating Pp and Ppk from specification limits and process spread using mu, sigma, and three-sigma bounds to illustrate process capability analysis and normal distribution.

Explore how to calculate and interpret Pp and Ppk for a sample, using x-bar and s to assess mean, spread, and alignment with upper and lower spec limits.

Learn to sample parts at regular intervals, compute x bar and range, build x bar charts, derive x double bar and R bar, and calculate Cp and Cpk.

Compute cp and cpk from sigma hat using R-bar and dx2 in SPC data, applying upper and lower spec limits with x-bar and control chart constants.

Master Cp and Cpk calculations in quality engineering statistics using Excel, with sample data, control chart constants, and interpretation of process capability.

Learn to interpret capability analysis by comparing identical data to varying specification limits and observe how P, P upper, P lower, CP, CP lower, PQ, and CPK respond.

Understand the difference between cpk and pq: cpk is the short-term process capability reflecting potential within subgroups, while pq measures long-term process performance across all data.

Summarize process capability analysis with an Excel tool, comparing 100% parts, batch samples, and repeated run samples, using mu, x-bar, x-bar double bar, sigma, and Cp/Cpk.

Apply process capability analysis and control charting through three practice problems with downloadable Excel workbooks and an answer key, drawn from real manufacturing scenarios and tied to design of experiments.

Join a four-factor, two-level designed experiment to improve tube cut length control, testing feed rule compound, feed speed, roll pressure, and die clamp pressure, measured by a length gauge.

See how a design of experiments isolates factor effects from four two-level variables, moving from confounded data to restored, ordered patterns.

Learn doe terminology by examining main effects, interactions, and experimental error; understand replication and blocking to separate factor influences from residual variation.

Plan a designed experiment to identify which factors influence process output and by how much, using a structured, team-driven approach with operators, maintenance, engineers, supervisors, and a facilitator.

Explore types of experimental designs, including latin square, jordan square, nest, and randomized block. Learn when to use factorial and fractional factorial designs to study variability, mean effects, and interactions.

Explore design of experiments concepts, including efficiency, orthogonal array, and the Taguchi lit matrix, and when to randomize, alongside measurement error factors such as accuracy, repeatability, linearity, reproducibility, and resolution.

Explore full factorial designs in engineering experiments, examining multiple factors across all treatment combinations and levels to reveal main and interactive effects on the average yield.

Examines full factorial and fractional factorial experiments to analyze how partition length, height, and gap affect air ventilation, using main effects, contrasts, ANOVA, and Taguchi designs.

Explore how fractional factorial designs use orthogonal arrays to improve efficiency over full factorial experiments, as Taguchi methods guide experiment setup, factor assignment, and interaction analysis.

Master design of experiments by working through nine practice problems in section h, using the data set, Excel workbook, and answer key to set up, execute, and interpret doe results.