Decision Making for Leaders: Leveraging Data

Explore how variation, driven by data, shapes leadership decisions, distinguishing normal from abnormal variation, and learn to measure and respond to it to improve quality and productivity.

Explore how leaders and managers use basic statistical methods to distinguish special and common cause variation and understand normal and non normal variation to guide decision making with suitable tools.

Recognize that variability is normal and predictable for groups, and use descriptive statistics, frequency distributions, and the normal distribution to anticipate and manage outcomes.

Identify the three sources of variability—normal variation around the mean, instability of the mean and width, and mean off target—and explain their impact on standard deviation and process targets.

Use a statistical process control chart to identify variability with the average and range charts, and view off target variability as a signal of potential improvements.

Identify out-of-control conditions using control chart rules: points beyond control limits, seven-point shifts and trends, and non-random patterns; act by investigating causes and applying corrective action.

Distinguish assignable and common cause variation, and empower leaders to reduce common cause variation by addressing five components: positional, part-to-part, batch-to-batch, process stream to process stream variation, and measurement error.

Assess measurement error by analyzing accuracy, precision, and linearity, then evaluate repeatability and reproducibility with MSA tools to reduce variability, cut costs, and improve quality.

Identify and separate sources of variation in the filled weight of shelf-stable cans, including within batch variation, between batch variation, and between filler seamer variation to optimize quality and cost.

Explore how the capability index (cpk) measures a process's ability to meet upper and lower specification limits, using an axles example with x-bar, sigma, and z minimum.

Understand how distributions shift, and reduce common and special cause variation through process adjustments, gauge improvements, tooling changes, material and supplier improvements, and management actions.

Improve process capability by analyzing Cpk, reducing variability, and shifting the center toward the lower spec limit under regulatory rules, addressing special and common cause variation.

Download the attached spreadsheet to practice data variation concepts with real formulas and raw data, and see how changing inputs affects outputs for practical learning at work.

Leaders learn to attack common cause variation using SPC charts, distinguish special vs common causes, and apply five sources of common cause variation to improve yield and CPK.

Explore how leaders use data and hypothesis testing to make rational decisions, quantify risk, and apply statistical tools without heavy math.

Learn to test null versus alternative hypotheses with a bun production case; mu equals 1000; determine if cheaper flour changes the mean output and whether to accept or reject H0.

Learn to use hypothesis testing to decide between alternatives and verify if groups or processes differ, enabling leaders to justify decisions with data, controlled tests, and error rates.

Define type one error (alpha) and type two error (beta) in null hypothesis testing; illustrate with the operating characteristic curve for accepting the proposal.

Explore how jury decisions illustrate type I and type II errors in decision making under uncertainty, weighing the risk of executing the innocent vs. releasing the guilty.

Explore sampling and distributions in manufacturing, focusing on the normal curve, center and width (standard deviation), and how 68% to 99.73% fall within ±1–3 standard deviations.

Explore measurement systems analysis to quantify gauge error and bias, and improve decisions by examining accuracy, repeatability, reproducibility, and linearity in gauged parts.

Compare z tests and t tests for means, noting z tests require known population standard deviation while t tests use sample standard deviation; for small samples, t tests are preferred.

Master t tests and z tests by hand or with commercial software such as Minitab and Key Macros. Enable the analysis toolpak in Excel for data analysis in hypothesis testing.

Apply the t statistic to test a supplier's claim that the population mean is 65, using 25 samples with a mean of 62.44 and standard deviation 3.056, at 5% risk.

Calculate the t statistic and compare it to the two-tailed critical value to decide whether to reject the null hypothesis; then apply these steps using Excel macros.

Perform a one-sample t test in Excel using a rubber hardness data set. Compute descriptive statistics, build a histogram, and decide on the null hypothesis with a two-tailed 5% test.

Conduct a one tailed t test to assess if a BMI sample differs from the population mean, using null hypothesis, 5% alpha, 19 df, and 1.729.

Demonstrate a longhand t-test on a BMI example with n=20, mu 23.5, x̄ 24.45, s 2.19, yielding t 1.94 against 1.729, and show Excel automation for the decision.

Explore how to run a one-tailed one-sample t-test for bmi data in excel using keymacro, comparing a sample mean to 23.5, and decide to reject the null hypothesis with p<0.05.

Explore t test concepts with Excel in practical examples, learn how to interpret results, and decide to accept or reject the null hypothesis across left-, right-, and two-tailed tests.

Assess whether a prescribed drug affects hemoglobin by conducting a two-tailed t-test on ten patients with alpha 0.05, using Excel; results reject the null and show an effect.

Explore how the feast and famine diet affects blood pressure using a paired t-test on eight patients. The cycle reduces blood pressure significantly, leading to rejection of the null hypothesis.

Compare writing scores between two freshman classes using one-tailed t tests for means, test the null hypothesis at 0.05, and benchmark teacher performance with Excel and KeyMacro analyses.

Compare welding students' performance when learning from an old book versus a new book using a paired two-sample t-test, showing a significant difference favoring the new book.

Use a t-test to assess whether delivery days average five days, compare the t statistic and p-value to a 5% alpha, and decide to accept the null hypothesis.

Evaluate whether cognitive training improves IQ using a sample of 15 students; a mean increase from 100 to 103.4 yields t = 3.329 and p = 0.002, rejecting the null.

compare fuel economy of radial tires versus bias ply tires using a two-tailed t test on twelve trials, reject the null as radial tires show slightly higher kilometers per liter.

Apply z tests for means when the population standard deviation is known; compare means with a two-tailed test using a breaking-strength example and a 1% alpha.

Explore z tests through Excel-based examples and a downloadable workbook. Work with raw data, input formulas, and practice hypothesis testing using the included worksheets.

Apply hypothesis testing to product strength data; in this fishing line example, the null is rejected by a z of -2.828, indicating lines do not meet the claimed strength.

Evaluate admissions testing improvements through a coaching program using a z-test on 25 students, with a sample mean of 536 versus a population mean of 500, indicating a positive effect.

A z test compares 42 applicants' mean score of 212.79 to the population mean of 210; with z=2.13 exceeding 1.645, we reject the null.

Assess a pharmaceutical claim by applying a two-tailed z test with population standard deviation 15, comparing treated group mean IQ to population mean, finding no significant difference (z = -0.94).

A one-tailed z test shows a mean score of 90 (n=81) surpassing the norm 82 with z=3.60, exceeding 1.645, and supporting the boast of superior performance.

Leverage z-tests in Excel to compare delivery times, using one- and two-tailed tests, with a 49-sample mean of 100 minutes against 120 minutes, alpha 0.05.

Conduct a two-tailed test at 5% using the z statistic to compare the average rod length to 50 cm; with 49.2 cm observed, reject the null due to length.

Test whether men's heart rate rises under stress from 71 bpm: nine participants show 73.9 bpm (sd 4), z=4.33, 5% one-tailed, rejecting stress does not change heart rate.

Assess whether technical school applicants have higher technical knowledge than the general population using a two-tailed z test at 5% alpha, but results show no significant difference.

Compare two centerless grinding machines using SPC range charts and z tests to decide if their mean outputs differ, concluding the means are effectively the same at 5% significance.

Develop data-driven decision making by practicing hypothesis testing with t tests and z tests through homework on anxiety, sleep, boiler emissions, nitrate effects on mice growth, and river alkalinity.

Practice hypothesis testing with an Excel workbook of five problems, using the provided data and Mike's notes to solidify understanding before moving to process metrics and the quality operating system.

Lead with a decision to improve by establishing a culture of continuous improvement and implementing the quality operating system.

Leaders learn to champion a continuous improvement culture through the quality operating system, setting aggressive yet realistic targets, and prioritizing urgent reviews, coaching, and process improvement.

Define what matters, measure it, and continually improve to create value for customers. Build the team, culture, and facility, deliver products and services, and reinvest profits to grow.

Discover the operating profit bridge and how wage and material cost inflation, plus price concessions, threaten profits; boost productivity with lean manufacturing and cost negotiation to turn revenue into profit.

Emphasize that continuous improvement pays off with 16 to 1 cost reductions, 6 to 1 revenue gains, and strong return on investment from process improvements.

Implement a quality operating system to manage important performance measurements with a data driven method, standardized tools and processes, supporting continuous improvement and driving quality, profit, and cash flow.

Learn how leaders leverage customer knowledge and data-driven decision making, pursue excellence through continuous improvement, identify waste, and plan to reach the top half.

Leaders drive data driven decision making across the whole organization, guiding continuous improvement with leading and lagging indicators, trend tracking, and 8D problem solving to achieve customer satisfaction.

Explore how leading indicators drive lagging outcomes in mental, physical, social, and financial well-being, using weight management as a concrete example of measuring targets, actuals, and corrective actions.

Explore lagging indicators across investors, company, and plants, including stock price, profit, cash growth, and quality, to guide leaders in selecting the most important metrics to track.

Leverage data-driven lagging indicators across sales, general, and administration to track costs, quality, inventory, and launches, guiding decisions on margins, supplier performance, and lean improvements.

Track inventory days against plan, target high-dollar components for reduction, and implement frequent milk runs, Pareto-based corrective actions, and quality improvements to boost overall plant performance.

Explore examples of leading and lagging indicators, showing how purchasing and product teams select key indicators and focus on leading indicators to influence process outcomes in a tool shop.

Introduce quality operating systems and decision making, emphasizing continuous improvement. Learn a ten-step process with leading and lagging indicators, plus templates and examples to empower leaders to improve.

Reflect on how data and decision-making techniques empower you to solve workplace problems and leverage lifetime access to course content.

Explore decision making for leaders leveraging data and related analytics courses, plus downloadable resources and a coupon link that shows the lowest price and supports career advancement.