Lean Six Sigma Green Belt · DMAIC · Updated 2026
Free Lean Six Sigma Green Belt Practice Exam (2026)
Full-length Green Belt mock exams with the same number of questions and time as common Green Belt certification exams. Define, Measure, Analyze, Improve and Control questions, answers explained, and your score by phase.
1. Choose a set
2. Choose a mode
100 questions · 3 hours · same as the real exam
You have an unfinished attempt on this set.
Questions 1–10Page 1 of 10
🏆
Question 1Define
Of 500 units that entered an assembly line, 440 passed on the first try, 40 failed but passed after rework, and 20 were scrapped. What is the first pass yield?
Correct answer: A. 88%
First pass yield counts only units that passed the first time without rework: 440 / 500 = 88%. The 96% figure is final yield (480 / 500), which hides the rework done in the hidden factory.
First pass yield counts only units that passed the first time without rework: 440 / 500 = 88%. The 96% figure is final yield (480 / 500), which hides the rework done in the hidden factory.
Question 2Measure
A line's takt time is 60 seconds, but the slowest station has a cycle time of 75 seconds. What is the MOST likely result if nothing changes?
Correct answer: C. The line will fall behind customer demand
When a station's cycle time is longer than takt time, the line cannot produce at the rate customers need, so backlogs or overtime follow. Takt time is set by demand and available time, so it does not change because a station is slow.
When a station's cycle time is longer than takt time, the line cannot produce at the rate customers need, so backlogs or overtime follow. Takt time is set by demand and available time, so it does not change because a station is slow.
Question 3Analyze
On a fishbone diagram for a paint defect, the team adds high workshop humidity as a possible cause. In the common 6M categories, where does it BEST fit?
Correct answer: D. Mother Nature (environment)
The environment category, often called Mother Nature, covers conditions such as humidity, temperature and dust. Machine covers equipment, not the surrounding conditions.
The environment category, often called Mother Nature, covers conditions such as humidity, temperature and dust. Machine covers equipment, not the surrounding conditions.
Question 4Improve
In a team's idea sessions, two senior engineers dominate and newer members rarely speak. Which technique BEST gives everyone an equal voice?
Correct answer: D. Nominal group technique
The nominal group technique has members generate ideas silently and then share them one at a time in turn, which limits domination by a few voices. Open brainstorming is what allowed the imbalance in the first place.
The nominal group technique has members generate ideas silently and then share them one at a time in turn, which limits domination by a few voices. Open brainstorming is what allowed the imbalance in the first place.
Question 5Control
A year after a project closed, the bakery replaces its proofing cabinet with a new model that has different settings. What should happen to the existing control plan?
Correct answer: B. Update it to reflect the new equipment
A control plan is a living document that must be revised whenever the process, equipment or requirements change. Leaving it unchanged means operators would monitor settings that no longer apply.
A control plan is a living document that must be revised whenever the process, equipment or requirements change. Leaving it unchanged means operators would monitor settings that no longer apply.
Question 6Analyze
A team wants to see whether variation in coating thickness comes mainly from position on a part, from part to part, or from hour to hour. Which tool is designed for this?
Correct answer: B. A multi-vari chart
A multi-vari chart displays variation by source, such as within a unit, between units and over time, to show which family of variation dominates. A Pareto chart ranks categories by count and does not separate these families.
A multi-vari chart displays variation by source, such as within a unit, between units and over time, to show which family of variation dominates. A Pareto chart ranks categories by count and does not separate these families.
Question 7Measure
On a standard process flowchart, what does a diamond-shaped symbol represent?
Correct answer: D. A decision point
A diamond marks a decision, usually a yes or no question with two exit paths. Start and end points are drawn as rounded rectangles or ovals.
A diamond marks a decision, usually a yes or no question with two exit paths. Start and end points are drawn as rounded rectangles or ovals.
Question 8Improve
A team generated 45 improvement ideas and needs to narrow them to a short list quickly, with each member getting a limited number of votes. Which technique is this?
Correct answer: C. Multivoting
Multivoting gives each member a fixed number of votes to spread across ideas, which quickly reduces a long list to a few favorites. Force field analysis looks at driving and restraining forces, not ranking ideas.
Multivoting gives each member a fixed number of votes to spread across ideas, which quickly reduces a long list to a few favorites. Force field analysis looks at driving and restraining forces, not ranking ideas.
Question 9Define
When calculating DPMO for a printed brochure, which item is MOST appropriate to count as a defect opportunity?
Correct answer: B. Each distinct feature the customer cares about
An opportunity is a distinct, measurable chance for a defect that matters to the customer, such as correct color or a correctly spelled price. Counting trivial items like pixels inflates the opportunity count and makes DPMO look artificially low.
An opportunity is a distinct, measurable chance for a defect that matters to the customer, such as correct color or a correctly spelled price. Counting trivial items like pixels inflates the opportunity count and makes DPMO look artificially low.
Question 10Control
A lab takes subgroups of 15 measurements at a time for a continuous quality characteristic. Which control chart is MOST appropriate?
Correct answer: C. Xbar-S chart
With larger subgroups, usually more than about 10, the standard deviation is a better estimate of spread than the range, so an Xbar-S chart is preferred. The np and c charts are for attribute data.
With larger subgroups, usually more than about 10, the standard deviation is a better estimate of spread than the range, so an Xbar-S chart is preferred. The np and c charts are for attribute data.
Question 11Measure
A hospital team wants to see how far nurses walk during a shift by tracing their actual paths on a floor plan. Which tool is BEST?
Correct answer: A. A spaghetti diagram
A spaghetti diagram traces the actual movement of people or materials on a layout, revealing wasted motion and transport. A value stream map shows flow and timing across the whole process, not walking paths.
A spaghetti diagram traces the actual movement of people or materials on a layout, revealing wasted motion and transport. A value stream map shows flow and timing across the whole process, not walking paths.
Question 12Analyze
A scatter plot of machine speed versus defect rate shows a clear U-shaped curve, but the correlation coefficient is close to zero. What is the BEST explanation?
Correct answer: B. The correlation coefficient measures only linear patterns
The correlation coefficient measures only the strength of a straight-line relationship, so a strong curved pattern can still give r near zero. This is why teams should always plot the data, not rely on r alone.
The correlation coefficient measures only the strength of a straight-line relationship, so a strong curved pattern can still give r near zero. This is why teams should always plot the data, not rely on r alone.
Question 13Define
A furniture retailer decides to add 100% final inspection to every order before shipping. Which cost of quality category increases DIRECTLY as a result?
Correct answer: A. Appraisal cost
Inspection and testing to find defects are appraisal costs. The goal may be to lower external failure costs, but the inspection itself is recorded as appraisal.
Inspection and testing to find defects are appraisal costs. The goal may be to lower external failure costs, but the inspection itself is recorded as appraisal.
Question 14Improve
A new labeling machine costs $30,000 and is expected to save $2,500 per month in rework. What is the simple payback period?
Correct answer: C. 12 months
Simple payback is cost divided by monthly savings: $30,000 / $2,500 = 12 months. Any shorter figure would require larger monthly savings than the estimate.
Simple payback is cost divided by monthly savings: $30,000 / $2,500 = 12 months. Any shorter figure would require larger monthly savings than the estimate.
Question 15Control
A packaging plant inspects exactly 200 cartons every shift and records the number of defective cartons. Which control chart is BEST?
Correct answer: D. np chart
An np chart tracks the number of defective units when the sample size is constant. A u chart tracks defects per unit, which counts each flaw rather than each defective carton.
An np chart tracks the number of defective units when the sample size is constant. A u chart tracks defects per unit, which counts each flaw rather than each defective carton.
Question 16Measure
Patients rate their satisfaction as very poor, poor, fair, good or excellent. What type of data is this?
Correct answer: D. Ordinal data
Ordinal data has categories with a natural order, but the gaps between them are not necessarily equal. Nominal data has categories with no order, such as product color.
Ordinal data has categories with a natural order, but the gaps between them are not necessarily equal. Nominal data has categories with no order, such as product color.
Question 17Analyze
A simple regression of scrap rate on line speed has an R-squared of 0.81 and a negative slope. What is the correlation coefficient?
Correct answer: A. -0.90
The correlation coefficient is the square root of R-squared, which is 0.90, and it takes the sign of the slope. With a negative slope, r = -0.90.
The correlation coefficient is the square root of R-squared, which is 0.90, and it takes the sign of the slope. With a negative slope, r = -0.90.
Question 18Define
A Six Sigma team at a call center plans changes to the call routing rules. Which role is accountable for the routing process day to day and should approve changes to how it runs?
Correct answer: B. The process owner
The process owner has authority over the process and is accountable for its daily performance, so changes need that person's agreement. The Green Belt leads the project but does not own the process.
The process owner has authority over the process and is accountable for its daily performance, so changes need that person's agreement. The Green Belt leads the project but does not own the process.
Question 19Improve
An experiment tests oven temperatures of 150 and 170 degrees using resin from two different suppliers. How many factors and levels does it have?
Correct answer: D. Two factors, each at two levels
The factors are temperature and supplier, and each is tested at two settings, so there are two factors at two levels. The four combinations are runs, not factors or levels.
The factors are temperature and supplier, and each is tested at two settings, so there are two factors at two levels. The four combinations are runs, not factors or levels.
Question 20Control
A software team tracks coding errors per 1,000 lines of code, and the size of each release varies widely. Which control chart is MOST appropriate?
Correct answer: C. u chart
A u chart tracks defects per unit when the amount of opportunity, such as lines of code, varies from sample to sample. A c chart assumes a constant area of opportunity.
A u chart tracks defects per unit when the amount of opportunity, such as lines of code, varies from sample to sample. A c chart assumes a constant area of opportunity.
Question 21Measure
A filler has 10 heads that fill bottles in rotation. A technician samples every 10th bottle for weight checks. What is the MAIN risk of this plan?
Correct answer: A. Every sample may come from the same head
Systematic sampling can be biased when the interval matches a cycle in the process, so every 10th bottle may always come from one head. A random start or a different interval would avoid that pattern.
Systematic sampling can be biased when the interval matches a cycle in the process, so every 10th bottle may always come from one head. A random start or a different interval would avoid that pattern.
Question 22Analyze
A model for repair cost is Cost = 500 + 25 x Hours. How should the 25 BEST be interpreted?
Correct answer: C. Each extra hour adds $25 to the predicted cost
The slope is the change in the predicted response for a one-unit increase in the predictor, so each extra hour adds $25. The $500 intercept is the predicted cost at zero hours, not the slope.
The slope is the change in the predicted response for a one-unit increase in the predictor, so each extra hour adds $25. The $500 intercept is the predicted cost at zero hours, not the slope.
Question 23Define
Which description BEST fits the typical role of a Green Belt?
Correct answer: A. Leads smaller projects part time alongside a regular job
Green Belts usually lead smaller-scope projects or support Black Belt projects while keeping their regular duties. Full-time leadership of large projects is the usual role of a Black Belt.
Green Belts usually lead smaller-scope projects or support Black Belt projects while keeping their regular duties. Full-time leadership of large projects is the usual role of a Black Belt.
Question 24Improve
A Green Belt plans to test three process settings by changing one factor at a time while holding the others fixed. Compared with a factorial design, what is the MAIN weakness of this approach?
Correct answer: B. It cannot detect interactions between factors
Changing one factor at a time cannot reveal interactions, and it is usually less efficient than a factorial design. A factorial design changes factors together so interactions can be estimated.
Changing one factor at a time cannot reveal interactions, and it is usually less efficient than a factorial design. A factorial design changes factors together so interactions can be estimated.
Question 25Control
On a control chart for drill hole depth, six consecutive points each rise above the one before, all within the control limits. What is the MOST likely cause to investigate?
Correct answer: A. Gradual tool wear
A steady run of increasing points is a trend signal, which often points to a gradual cause such as tool wear or rising temperature. A single typing error would usually produce one odd point, not a trend.
A steady run of increasing points is a trend signal, which often points to a gradual cause such as tool wear or rising temperature. A single typing error would usually produce one odd point, not a trend.
Question 26Measure
A data set of repair times has a mean of 42 minutes and a median of 30 minutes. What is the MOST likely shape of the distribution?
Correct answer: C. Skewed to the right
When the mean is noticeably higher than the median, a long tail of large values is pulling the mean up, which is a right or positive skew. In a symmetric distribution the mean and median are about equal.
When the mean is noticeably higher than the median, a long tail of large values is pulling the mean up, which is a right or positive skew. In a symmetric distribution the mean and median are about equal.
Question 27Analyze
A team believes a new scheduling method will REDUCE mean patient wait time compared with the old method. Which alternative hypothesis fits this belief?
Correct answer: B. Mean new wait is less than mean old wait
The alternative hypothesis states the effect the team wants evidence for, here that the new mean is lower, which makes it a one-sided test. The statement of equality is the null hypothesis.
The alternative hypothesis states the effect the team wants evidence for, here that the new mean is lower, which makes it a one-sided test. The statement of equality is the null hypothesis.
Question 28Define
A Champion asks a Green Belt which part of the project charter explains why the project matters to the company now, including its link to strategy and expected financial impact. Which section is this?
Correct answer: D. The business case
The business case explains why the project is worth doing, linking it to strategy and expected benefits. The problem statement describes the current gap with data but not the strategic justification.
The business case explains why the project is worth doing, linking it to strategy and expected benefits. The problem statement describes the current gap with data but not the strategic justification.
Question 29Improve
In a designed experiment, why does a team replicate some or all of the runs?
Correct answer: A. To estimate the experimental error
Replication repeats runs under the same settings, which gives an estimate of experimental error for judging whether effects are real. It does not replace randomization, which protects against hidden time effects.
Replication repeats runs under the same settings, which gives an estimate of experimental error for judging whether effects are real. It does not replace randomization, which protects against hidden time effects.
Question 30Control
For 15 consecutive subgroups, every point on an Xbar chart falls within one standard deviation of the center line. What should the team suspect?
Correct answer: B. The subgroups mix streams or the limits are wrong
Points hugging the center line for a long run is a recognized signal, often caused by stratified subgroups that mix several streams or by incorrectly calculated limits. It should be investigated rather than celebrated.
Points hugging the center line for a long run is a recognized signal, often caused by stratified subgroups that mix several streams or by incorrectly calculated limits. It should be investigated rather than celebrated.
Question 31Measure
Cycle times have a mean of 200 seconds and a standard deviation of 8 seconds. The upper specification limit is 224 seconds. What is the Z value for the upper limit?
Correct answer: A. 3.0
Z = (USL - mean) / standard deviation = (224 - 200) / 8 = 3.0. The value 24 is the raw distance in seconds before dividing by the standard deviation.
Z = (USL - mean) / standard deviation = (224 - 200) / 8 = 3.0. The value 24 is the raw distance in seconds before dividing by the standard deviation.
Question 32Analyze
A team wants to compare the typical resolution times of two support teams. The data are heavily skewed with several extreme values, and each group has only 14 observations. Which test is MOST appropriate?
Correct answer: D. Mann-Whitney test comparing medians
A nonparametric test such as Mann-Whitney compares medians without assuming normal data, which suits small, heavily skewed samples. A 2-sample t-test assumes roughly normal data, which these small skewed samples do not support.
A nonparametric test such as Mann-Whitney compares medians without assuming normal data, which suits small, heavily skewed samples. A 2-sample t-test assumes roughly normal data, which these small skewed samples do not support.
Question 33Define
A team is building a SIPOC for a mortgage approval process. A credit bureau sends a credit report that the underwriters use. How should the credit bureau be listed?
Correct answer: A. As a supplier
In a SIPOC, the supplier is whoever provides an input, so the credit bureau is the supplier and the credit report is the input. Listing the bureau itself as an input confuses the provider with what it provides.
In a SIPOC, the supplier is whoever provides an input, so the credit bureau is the supplier and the credit report is the input. Listing the bureau itself as an input confuses the provider with what it provides.
Question 34Improve
A team has 9 possible factors that might affect weld strength and a limited budget for experiments. What is the MOST efficient first step?
Correct answer: C. A fractional factorial screening design
A fractional factorial screening design tests many factors with far fewer runs to find the few that matter. A full factorial with 9 factors at two levels would need 512 runs.
A fractional factorial screening design tests many factors with far fewer runs to find the few that matter. A full factorial with 9 factors at two levels would need 512 runs.
Question 35Control
Over 25 days, a clinic checked 5,000 patient records in total and found 150 with missing data. What is the center line of the p chart?
Correct answer: A. 0.03
The center line is total defectives divided by total inspected: 150 / 5,000 = 0.03, or 3%. The 0.06 figure comes from dividing 150 by 25 days and then by 100, which mixes up units.
The center line is total defectives divided by total inspected: 150 / 5,000 = 0.03, or 3%. The 0.06 figure comes from dividing 150 by 25 days and then by 100, which mixes up units.
Question 36Measure
A box plot of call times has a first quartile of 20 minutes and a third quartile of 30 minutes. Using the 1.5 x IQR rule, which call time would be flagged as an outlier?
Correct answer: D. 47 minutes
The IQR is 30 - 20 = 10, so the upper fence is 30 + 1.5 x 10 = 45 minutes. Only 47 minutes is above the fence, so 44 minutes is within the whisker range.
The IQR is 30 - 20 = 10, so the upper fence is 30 + 1.5 x 10 = 45 minutes. Only 47 minutes is above the fence, so 44 minutes is within the whisker range.
Question 37Analyze
A pharmaceutical team tests a new mixing method and finds no significant difference in blend uniformity, so they drop the idea. In reality, the new method was better. Which error occurred?
Correct answer: B. A Type II error
A Type II error is failing to reject the null hypothesis when the alternative is true, so a real improvement is missed. An alpha error is another name for a Type I error, which is a false alarm.
A Type II error is failing to reject the null hypothesis when the alternative is true, so a real improvement is missed. An alpha error is another name for a Type I error, which is a false alarm.
Question 38Define
A kitchen appliance maker wants to learn how home cooks actually use its blenders, including steps users may not think to mention. Which voice of the customer method is BEST?
Correct answer: C. Watching cooks use it at home
Direct observation in the customer's real setting, often called a gemba visit, reveals behaviors and pain points that people do not report. A survey only captures what customers think to say, so unspoken needs are missed.
Direct observation in the customer's real setting, often called a gemba visit, reveals behaviors and pain points that people do not report. A survey only captures what customers think to say, so unspoken needs are missed.
Question 39Improve
A pilot of a new intake process cut wait times, but by less than the project goal, and staff found two steps confusing. What should the team do NEXT?
Correct answer: B. Refine the solution and pilot again
A pilot exists to learn, so the team should study what fell short, adjust the solution and test again before a full rollout. Lowering the goal to fit the result ignores what the pilot revealed.
A pilot exists to learn, so the team should study what fell short, adjust the solution and test again before a full rollout. Lowering the goal to fit the result ignores what the pilot revealed.
Question 40Control
A c chart for weekly safety incidents has an average count of 9. What is the upper control limit?
Correct answer: D. 18
For a c chart, UCL = c-bar + 3 x square root of c-bar = 9 + 3 x 3 = 18. The 12 answer adds only one standard deviation of 3.
For a c chart, UCL = c-bar + 3 x square root of c-bar = 9 + 3 x 3 = 18. The 12 answer adds only one standard deviation of 3.
Question 41Measure
A team plotted a month of hourly temperature readings as a histogram, and it looks roughly normal. Operators still suspect the temperature drifts upward during each day. What should the team do NEXT?
Correct answer: B. Plot the data in time order on a run chart
A histogram hides time order, so it cannot show drifts or cycles. A run chart plots the data in sequence and would reveal a daily upward pattern.
A histogram hides time order, so it cannot show drifts or cycles. A run chart plots the data in sequence and would reveal a daily upward pattern.
Question 42Analyze
A Green Belt worries that her hypothesis test may miss a real difference of practical size. Which action MOST directly increases the power of the test?
Correct answer: C. Increasing the sample size
Power is the chance of detecting a real difference, and a larger sample increases it. Lowering alpha makes a Type I error less likely but reduces power, all else being equal.
Power is the chance of detecting a real difference, and a larger sample increases it. Lowering alpha makes a Type I error less likely but reduces power, all else being equal.
Question 43Define
A software firm collected 300 open-ended customer comments and wants to group them into natural themes before defining CTQs. Which tool is BEST?
Correct answer: A. An affinity diagram
An affinity diagram groups large numbers of ideas or verbal comments into related themes. A Pareto chart ranks categories by frequency but needs the categories to exist first.
An affinity diagram groups large numbers of ideas or verbal comments into related themes. A Pareto chart ranks categories by frequency but needs the categories to exist first.
Question 44Improve
A plant builds three products, but it runs all of product A on Monday and Tuesday and product C on Friday, causing big swings in workload and inventory. Which Lean concept addresses this?
Correct answer: D. Heijunka (production leveling)
Heijunka levels the volume and mix of production over time, for example by building small amounts of each product every day. Jidoka is about stopping to fix problems, not leveling the schedule.
Heijunka levels the volume and mix of production over time, for example by building small amounts of each product every day. Jidoka is about stopping to fix problems, not leveling the schedule.
Question 45Control
While calculating initial control limits, a team finds two points caused by a documented power failure that has since been fixed. How should these points be treated?
Correct answer: A. Exclude them and recalculate the limits
Points with a known, removed special cause are usually excluded so the limits reflect the normal process. Keeping them would inflate the limits and hide future signals.
Points with a known, removed special cause are usually excluded so the limits reflect the normal process. Keeping them would inflate the limits and hide future signals.
Question 46Measure
In a gage R&R study, each of three operators measures the same parts with the same gauge. One operator's average readings are consistently higher than the others. Which component of measurement error does this indicate?
Correct answer: C. Reproducibility
Reproducibility, or appraiser variation, is the difference between operators measuring the same parts. Training or a standard method usually reduces it.
Reproducibility, or appraiser variation, is the difference between operators measuring the same parts. Training or a standard method usually reduces it.
Question 47Analyze
Fifteen technicians take a skills test before training and the same test after training. The team wants to know whether the mean score changed. Which test is MOST appropriate?
Correct answer: D. Paired t-test
A paired t-test analyzes the differences between two measurements on the same subjects, removing person-to-person variation. A 2-sample t-test assumes independent groups, which these are not.
A paired t-test analyzes the differences between two measurements on the same subjects, removing person-to-person variation. A 2-sample t-test assumes independent groups, which these are not.
Question 48Define
Operators on a packaging line have repeatedly suggested a simple fix for jams, but supervisors never ask for or act on their ideas. Which of the eight wastes does this show?
Correct answer: D. Non-utilized talent
Non-utilized talent is failing to use people's skills, knowledge and ideas. The jams themselves may cause waiting, but the waste described is ignoring the operators' input.
Non-utilized talent is failing to use people's skills, knowledge and ideas. The jams themselves may cause waiting, but the waste described is ignoring the operators' input.
Question 49Improve
Two processes cannot be directly linked in continuous flow because one runs in batches. What Lean tool controls the inventory between them so the downstream step pulls only what it needs?
Correct answer: B. A supermarket
A supermarket is a controlled store of parts between processes, replenished as the downstream process pulls items. It allows pull to work where true one-piece flow is not possible.
A supermarket is a controlled store of parts between processes, replenished as the downstream process pulls items. It allows pull to work where true one-piece flow is not possible.
Question 50Control
A control plan relies on a coating thickness gauge. What should the plan include to keep the data trustworthy over time?
Correct answer: C. Periodic calibration and gauge checks
Measurement systems can drift, so control plans include periodic calibration and repeat measurement system checks. A single study at the start does not protect against later drift.
Measurement systems can drift, so control plans include periodic calibration and repeat measurement system checks. A single study at the start does not protect against later drift.
Question 51Measure
A gage R&R study reports a number of distinct categories (ndc) of 2. What does this mean?
Correct answer: D. The gauge cannot tell parts apart well enough
The number of distinct categories shows how many groups the measurement system can reliably separate, and a value of 5 or more is usually wanted. A value of 2 means the gauge cannot discriminate well between parts.
The number of distinct categories shows how many groups the measurement system can reliably separate, and a value of 5 or more is usually wanted. A value of 2 means the gauge cannot discriminate well between parts.
Question 52Analyze
A dairy wants to know whether the mean weight of its yogurt cups differs from the labeled target of 500 grams, using a sample of 30 cups. Which test is MOST appropriate?
Correct answer: C. 1-sample t-test
A 1-sample t-test compares one group's mean to a known target or standard value. A 2-sample t-test needs two separate groups of data.
A 1-sample t-test compares one group's mean to a known target or standard value. A 2-sample t-test needs two separate groups of data.
Question 53Define
A plant keeps its stamping press running during slow weeks to keep the machine busy, building parts that have no customer orders yet. Which waste is this?
Correct answer: B. Overproduction
Overproduction is making more than needed or before it is needed, and it is often called the worst waste because it creates others. The stock that piles up is inventory, but the root waste is the decision to overproduce.
Overproduction is making more than needed or before it is needed, and it is often called the worst waste because it creates others. The stock that piles up is inventory, but the root waste is the decision to overproduce.
Question 54Improve
On an assembly line, any operator who spots a defect pulls a cord that lights a signal and can stop the line until the problem is fixed. Which Lean concept is this?
Correct answer: C. Jidoka with an andon
Jidoka means building quality in by stopping when a problem occurs, and an andon is the signal that calls for help. Kanban signals replenishment, not quality problems.
Jidoka means building quality in by stopping when a problem occurs, and an andon is the signal that calls for help. Kanban signals replenishment, not quality problems.
Question 55Control
A project at one distribution center cut picking errors by 60%. The company has eight similar centers. What is the BEST next step to maximize benefit?
Correct answer: A. Replicate the solution at similar sites
Replicating, or translating, a proven solution to similar processes multiplies the benefit with far less effort than new projects. Each site should still adapt and verify the solution for local differences.
Replicating, or translating, a proven solution to similar processes multiplies the benefit with far less effort than new projects. Each site should still adapt and verify the solution for local differences.
Question 56Measure
A lab measures the same certified reference block every Monday for three months. The readings drift slowly upward over time. Which measurement system property is the problem?
Correct answer: B. Stability
Stability is the consistency of a measurement system's readings on the same standard over time, so a slow drift is a stability problem. Linearity concerns how bias changes across the measuring range.
Stability is the consistency of a measurement system's readings on the same standard over time, so a slow drift is a stability problem. Linearity concerns how bias changes across the measuring range.
Question 57Analyze
A one-way ANOVA comparing mean cycle time across five work cells gives a p-value of 0.003. What should the team do NEXT to find out which cells differ?
Correct answer: A. Run pairwise comparisons such as Tukey
A significant ANOVA shows that at least one mean differs, but not which ones. Follow-up multiple comparison methods, such as Tukey's test, identify the specific pairs that differ.
A significant ANOVA shows that at least one mean differs, but not which ones. Follow-up multiple comparison methods, such as Tukey's test, identify the specific pairs that differ.
Question 58Define
A repair shop sorted, labeled and cleaned its tool area. Three months later, tools are back on benches and labels are missing. Which 5S step was MOST likely neglected?
Correct answer: C. Sustain
Sustain is the discipline of keeping the standards through audits, habits and leadership attention. The earlier steps were done, but without sustain the area drifted back to its old state.
Sustain is the discipline of keeping the standards through audits, habits and leadership attention. The earlier steps were done, but without sustain the area drifted back to its old state.
Question 59Improve
A team rearranges five scattered machines into a U-shaped cell so one operator can move parts directly from step to step. What is the MAIN benefit?
Correct answer: D. Better flow with less transport and WIP
A U-shaped cell supports one-piece flow, cutting transport, waiting and work in process. Larger batches and more storage are the opposite of what cellular layouts aim for.
A U-shaped cell supports one-piece flow, cutting transport, waiting and work in process. Larger batches and more storage are the opposite of what cellular layouts aim for.
Question 60Control
What is the MAIN purpose of statistical process control in the Control phase?
Correct answer: D. To detect special causes quickly over time
Statistical process control monitors a process over time to tell common cause from special cause variation, so problems are caught early. It does not set specifications, which come from the customer.
Statistical process control monitors a process over time to tell common cause from special cause variation, so problems are caught early. It does not set specifications, which come from the customer.
Question 61Measure
A process has a mean of 53, a standard deviation of 1, a lower specification of 46 and an upper specification of 56. What is Cpk?
Correct answer: A. 1.00
Cpu = (56 - 53) / (3 x 1) = 1.00 and Cpl = (53 - 46) / (3 x 1) = 2.33. Cpk is the smaller of the two, 1.00. The value 1.67 is Cp, which ignores how far the mean sits from center.
Cpu = (56 - 53) / (3 x 1) = 1.00 and Cpl = (53 - 46) / (3 x 1) = 2.33. Cpk is the smaller of the two, 1.00. The value 1.67 is Cp, which ignores how far the mean sits from center.
Question 62Analyze
A purchasing team wants to compare the percentage of defective shipments from two suppliers, using pass or fail results for each shipment. Which test is MOST appropriate?
Correct answer: B. 2-proportion test
A 2-proportion test compares the proportions of a pass or fail outcome between two groups. A 2-sample t-test compares means of continuous data, not proportions.
A 2-proportion test compares the proportions of a pass or fail outcome between two groups. A 2-sample t-test compares means of continuous data, not proportions.
Question 63Define
A pharmacy must record a second pharmacist's check on every controlled drug order because regulations require it. Customers do not value this step directly. How is it BEST classified?
Correct answer: D. Business non-value-added work
Steps required by law or the business that customers do not pay for are business non-value-added. They should be kept but done as efficiently as possible, not eliminated outright.
Steps required by law or the business that customers do not pay for are business non-value-added. They should be kept but done as efficiently as possible, not eliminated outright.
Question 64Improve
In an FMEA, a team gives a failure mode a detection rating of 2. What does this rating mean?
Correct answer: C. The failure is very likely to be detected
On the usual 1 to 10 scale, a low detection rating means current controls are very likely to catch the failure before it reaches the customer. A rating near 10 means the failure would almost never be detected.
On the usual 1 to 10 scale, a low detection rating means current controls are very likely to catch the failure before it reaches the customer. A rating near 10 means the failure would almost never be detected.
Question 65Control
A process owner hires three new operators after a project closes. What is the BEST way to make sure they follow the improved method?
Correct answer: B. Train them on the revised SOP and record it
Formal training on the current SOP, with records such as a skills matrix, builds the new method into onboarding. Informal shadowing can pass on old habits and shortcuts.
Formal training on the current SOP, with records such as a skills matrix, builds the new method into onboarding. Informal shadowing can pass on old habits and shortcuts.
Question 66Measure
A customer asks a supplier to show its process is capable. Which Cpk value is MOST often used as the minimum for an existing process?
Correct answer: C. 1.33
A Cpk of 1.33 is a widely used minimum for existing processes, giving a margin of about 4 sigma to the nearest limit. A Cpk of 1.00 means the process just fits, with no margin for shifts.
A Cpk of 1.33 is a widely used minimum for existing processes, giving a margin of about 4 sigma to the nearest limit. A Cpk of 1.00 means the process just fits, with no margin for shifts.
Question 67Analyze
A process shows only common cause variation on its control chart, but the spread is still too wide for customers. What is MOST likely needed to reduce the variation?
Correct answer: A. A fundamental change to the process itself
Common cause variation comes from the system itself, so reducing it requires changing the process, such as its methods, materials or equipment. Reacting to individual points treats common cause as special cause and adds variation.
Common cause variation comes from the system itself, so reducing it requires changing the process, such as its methods, materials or equipment. Reacting to individual points treats common cause as special cause and adds variation.
Question 68Define
An online retailer processes 1,200 returns per year caused by wrong items being shipped. Each return costs $35 in handling plus $15 in shipping. What is the annual cost of poor quality from these returns?
Correct answer: B. $60,000
Each return costs $35 + $15 = $50, so 1,200 x $50 = $60,000 per year. The $42,000 figure includes only the handling cost and leaves out shipping.
Each return costs $35 + $15 = $50, so 1,200 x $50 = $60,000 per year. The $42,000 figure includes only the handling cost and leaves out shipping.
Question 69Improve
A team has chosen a solution that changes how orders are approved. Before implementing it, the Green Belt wants to anticipate how the new process could fail and add safeguards. Which tool is BEST?
Correct answer: A. An FMEA of the new process
An FMEA applied to the proposed process identifies new failure modes, rates their risk and drives preventive actions before rollout. A Pareto of past complaints looks backward at the old process, not risks in the new one.
An FMEA applied to the proposed process identifies new failure modes, rates their risk and drives preventive actions before rollout. A Pareto of past complaints looks backward at the old process, not risks in the new one.
Question 70Control
At the Control phase tollgate, what is the Champion MOST likely to confirm before closing the project?
Correct answer: C. Results hold and the owner accepts the plan
The Control tollgate checks that improvements are holding, controls are in place and the process owner has accepted ownership. Baseline data collection belongs in Measure, not at project closure.
The Control tollgate checks that improvements are holding, controls are in place and the process owner has accepted ownership. Baseline data collection belongs in Measure, not at project closure.
Question 71Measure
A team collected 500 cycle times but did not record which line, shift or operator produced each one. Later they cannot tell where the delays come from. What should the data collection plan have included?
Correct answer: B. Stratification factors for each record
Recording stratification factors such as line, shift and operator with each data point allows the team to slice the data later. A larger sample without those factors would still not show where the delays occur.
Recording stratification factors such as line, shift and operator with each data point allows the team to slice the data later. A larger sample without those factors would still not show where the delays occur.
Question 72Analyze
A 95% confidence interval for the difference in mean cycle time between two methods runs from -1.2 to 3.5 seconds. What does this suggest?
Correct answer: D. There is no significant difference at the 5% level
Because the interval includes zero, a difference of zero is plausible, so the result is not significant at alpha 0.05. If the whole interval were above or below zero, the difference would be significant.
Because the interval includes zero, a difference of zero is plausible, so the result is not significant at alpha 0.05. If the whole interval were above or below zero, the difference would be significant.
Question 73Define
Members of an HR project team are unclear about who does each task, who must approve it and who only needs to be kept informed. Which tool BEST resolves this?
Correct answer: A. A RACI chart
A RACI chart shows who is Responsible, Accountable, Consulted and Informed for each task or decision. A Gantt chart shows timing and sequence, not decision rights.
A RACI chart shows who is Responsible, Accountable, Consulted and Informed for each task or decision. A Gantt chart shows timing and sequence, not decision rights.
Question 74Improve
After mapping the current state of an order process, a team draws a map of how the process should look once waste is removed and flow is improved. What is this map used for?
Correct answer: B. Setting the target for improvement
A future-state value stream map shows the target design and guides the improvement plan. It does not replace the charter or serve as a record of past performance.
A future-state value stream map shows the target design and guides the improvement plan. It does not replace the charter or serve as a record of past performance.
Question 75Control
A Green Belt finishes a project and prepares a concise storyboard of the problem, analysis, solutions and results. What is the MAIN value of this documentation?
Correct answer: D. It shares knowledge with others in the company
Project documentation lets others learn from the work, supports audits and makes replication easier. It does not replace updating SOPs, which is what actually guides daily work.
Project documentation lets others learn from the work, supports audits and makes replication easier. It does not replace updating SOPs, which is what actually guides daily work.
Question 76Measure
A Green Belt is planning sample sizes. Compared with continuous data, what is generally true of attribute data such as pass or fail?
Correct answer: D. It needs larger samples for the same precision
Attribute data carries less information per observation than continuous data, so larger samples are needed for the same precision. This is one reason teams prefer continuous measures when they are available.
Attribute data carries less information per observation than continuous data, so larger samples are needed for the same precision. This is one reason teams prefer continuous measures when they are available.
Question 77Analyze
A Green Belt's goal is to reduce the spread of fill volumes, and she wants to know whether a new nozzle has less variation than the old one. Which test is MOST appropriate?
Correct answer: A. A test for equal variances
A test for equal variances, such as an F-test or Levene's test, compares the spread of two groups. A t-test compares means, which would not answer a question about variation.
A test for equal variances, such as an F-test or Levene's test, compares the spread of two groups. A t-test compares means, which would not answer a question about variation.
Question 78Define
At the end of the Define phase, the team meets with the Champion to review the charter, SIPOC and VOC findings. What is the MAIN purpose of this tollgate review?
Correct answer: C. To decide whether the project proceeds to Measure
A tollgate review checks that the phase deliverables are complete and decides whether the project should continue, change or stop. Choosing a solution happens in Improve, after the root causes are known.
A tollgate review checks that the phase deliverables are complete and decides whether the project should continue, change or stop. Choosing a solution happens in Improve, after the root causes are known.
Question 79Improve
A team reviews each step of an expense approval process asking whether it can be eliminated, combined with another, rearranged or simplified. Which approach is this?
Correct answer: A. ECRS
ECRS stands for eliminate, combine, rearrange and simplify, and it is a simple way to redesign process steps. RACI clarifies roles, which is a different purpose.
ECRS stands for eliminate, combine, rearrange and simplify, and it is a simple way to redesign process steps. RACI clarifies roles, which is a different purpose.
Question 80Control
A process owner worries that a standard Shewhart chart will be slow to detect small, gradual shifts of about 1 standard deviation. Which chart type is designed to be more sensitive to such shifts?
Correct answer: B. An EWMA or CUSUM chart
EWMA and CUSUM charts accumulate information from past points, which makes them better at detecting small sustained shifts. A Pareto chart and a histogram do not track a process over time.
EWMA and CUSUM charts accumulate information from past points, which makes them better at detecting small sustained shifts. A Pareto chart and a histogram do not track a process over time.
Question 81Measure
Which statement about a normal distribution is TRUE?
Correct answer: B. The mean, median and mode are equal
A normal distribution is symmetric and bell-shaped, so its mean, median and mode are all equal. Its tails extend indefinitely, although very little data falls beyond plus or minus 3 sigma.
A normal distribution is symmetric and bell-shaped, so its mean, median and mode are all equal. Its tails extend indefinitely, although very little data falls beyond plus or minus 3 sigma.
Question 82Analyze
A regression of fuel use on load weight looks strong, but one data point lies far from all the others and pulls the line toward it. What should the team do FIRST?
Correct answer: C. Investigate why that point is so different
Unusual points should be investigated to see if they are errors or real behavior before any decision is made. Deleting a point simply because it does not fit can hide important information about the process.
Unusual points should be investigated to see if they are errors or real behavior before any decision is made. Deleting a point simply because it does not fit can hide important information about the process.
Question 83Define
A manager proposes four improvement ideas. Which one is the BEST fit for a full DMAIC project?
Correct answer: B. Cutting a 9% billing error rate with unknown causes
DMAIC suits problems that are measurable, important and whose root cause is not known. Problems with an obvious, known fix should simply be implemented without a full project.
DMAIC suits problems that are measurable, important and whose root cause is not known. Problems with an obvious, known fix should simply be implemented without a full project.
Question 84Improve
A distribution center team studies how a top-performing company in another industry handles fast order picking, to adapt its ideas. What is this practice called?
Correct answer: D. Benchmarking
Benchmarking compares processes with leading practices, often in other industries, to find ideas for improvement. Stratification is a data analysis technique, not a source of solutions.
Benchmarking compares processes with leading practices, often in other industries, to find ideas for improvement. Stratification is a data analysis technique, not a source of solutions.
Question 85Control
A project goal was to cut a defect rate by at least 70%. The baseline rate was 8%, and after improvements the stable rate is 2%. Has the goal been met?
Correct answer: A. Yes, the reduction is 75%
The relative reduction is (8% - 2%) / 8% = 6 / 8 = 75%, which beats the 70% goal. The 6% figure is the drop in percentage points, not the relative reduction.
The relative reduction is (8% - 2%) / 8% = 6 / 8 = 75%, which beats the 70% goal. The 6% figure is the drop in percentage points, not the relative reduction.
Question 86Measure
A capability report shows both Cpk and Ppk for a molding process. What is the key difference between them?
Correct answer: C. Ppk uses overall variation over time
Ppk uses the overall, long-term standard deviation, while Cpk uses the within-subgroup, short-term estimate. Because overall variation includes shifts between subgroups, Ppk is usually equal to or lower than Cpk.
Ppk uses the overall, long-term standard deviation, while Cpk uses the within-subgroup, short-term estimate. Because overall variation includes shifts between subgroups, Ppk is usually equal to or lower than Cpk.
Question 87Analyze
In an assembly line, work-in-process piles up in front of the soldering station, while the stations after it are often idle. What does this MOST likely indicate?
Correct answer: A. The soldering station is the bottleneck
Inventory building up before a step and starved steps after it are classic signs of a bottleneck. Improving any other station would not raise the line's output until the bottleneck is addressed.
Inventory building up before a step and starved steps after it are classic signs of a bottleneck. Improving any other station would not raise the line's output until the bottleneck is addressed.
Question 88Define
A process runs at about 6,210 defects per million opportunities. Using the conventional 1.5 sigma shift, what is its approximate sigma level?
Correct answer: D. 4 sigma
With the standard 1.5 sigma shift, 6,210 DPMO corresponds to about 4 sigma. A 3 sigma process runs at about 66,807 DPMO, which is far worse.
With the standard 1.5 sigma shift, 6,210 DPMO corresponds to about 4 sigma. A 3 sigma process runs at about 66,807 DPMO, which is far worse.
Question 89Improve
A pilot of a new invoicing method ran for four weeks. How should the team BEST confirm that the improvement is real and not just chance?
Correct answer: A. Test pilot data against the baseline
A hypothesis test comparing pilot results with baseline data shows whether the change is statistically significant. Opinions or a single best day can be misleading and do not account for normal variation.
A hypothesis test comparing pilot results with baseline data shows whether the change is statistically significant. Opinions or a single best day can be misleading and do not account for normal variation.
Question 90Control
A Green Belt project has closed with strong results. What is a good practice for recognizing the team?
Correct answer: C. Celebrate and thank the team publicly
Recognizing the whole team reinforces the improvement culture and encourages future participation. Rewarding only one person can discourage the cooperation that projects rely on.
Recognizing the whole team reinforces the improvement culture and encourages future participation. Rewarding only one person can discourage the cooperation that projects rely on.
Question 91Measure
A team maps a purchase approval process using only the written procedure. Staff say the real process involves extra steps and workarounds. What should the team do?
Correct answer: D. Walk the process to map how it really runs
Measure needs an as-is map of how the work actually happens, which means going to the gemba and observing the real steps. A map of the written procedure would hide the workarounds that often cause problems.
Measure needs an as-is map of how the work actually happens, which means going to the gemba and observing the real steps. A map of the written procedure would hide the workarounds that often cause problems.
Question 92Analyze
A team plans a 2-sample t-test using only 12 measurements in each group. What assumption should they check FIRST?
Correct answer: D. That each group is roughly normally distributed
With small samples, t-tests assume the data in each group are roughly normal, which can be checked with a normal probability plot. Categorical counts call for a chi-square test instead.
With small samples, t-tests assume the data in each group are roughly normal, which can be checked with a normal probability plot. Categorical counts call for a chi-square test instead.
Question 93Define
Lean thinking is often summarized in five principles: identify value, map the value stream, create flow, establish pull and pursue perfection. Which principle comes right after mapping the value stream?
Correct answer: B. Create flow
After value is defined and the value stream is mapped, the next step is to make the remaining steps flow without interruption. Pull is introduced after flow, so production is triggered by customer demand.
After value is defined and the value stream is mapped, the next step is to make the remaining steps flow without interruption. Pull is introduced after flow, so production is triggered by customer demand.
Question 94Improve
A team scores five solutions. One scores highest overall but breaks a mandatory data privacy regulation. How should the team treat it?
Correct answer: C. Eliminate it, since it fails a must criterion
Must-have criteria, such as legal compliance, are applied as screens before weighted scoring, and any option that fails them is removed. A high score cannot make up for breaking a mandatory requirement.
Must-have criteria, such as legal compliance, are applied as screens before weighted scoring, and any option that fails them is removed. A high score cannot make up for breaking a mandatory requirement.
Question 95Control
A control chart uses limits set at plus or minus 3 standard deviations from the center line. For a stable, normally distributed process, about how often will a single point fall outside the limits by chance?
Correct answer: B. About 0.27% of points
About 99.73% of a normal distribution falls within plus or minus 3 sigma, so roughly 0.27% of points fall outside by chance. The 5% figure matches plus or minus 2 sigma, which would cause many more false alarms.
About 99.73% of a normal distribution falls within plus or minus 3 sigma, so roughly 0.27% of points fall outside by chance. The 5% figure matches plus or minus 2 sigma, which would cause many more false alarms.
Question 96Measure
Over 10 working days, a warehouse had 2 picking errors on each of 6 days and 7 picking errors on each of the other 4 days. What is the mean number of errors per day?
Correct answer: C. 4.0
Total errors are 6 x 2 + 4 x 7 = 12 + 28 = 40, and 40 / 10 days = 4.0 errors per day. The 4.5 answer simply averages 2 and 7 without weighting by the number of days.
Total errors are 6 x 2 + 4 x 7 = 12 + 28 = 40, and 40 / 10 days = 4.0 errors per day. The 4.5 answer simply averages 2 and 7 without weighting by the number of days.
Question 97Analyze
A plant made 500 parts: 200 on the night shift and 300 on the day shift. In total, 50 parts were defective. For a chi-square test, what is the expected number of defective night shift parts if shift and defects are unrelated?
Correct answer: B. 20
Expected count = row total x column total / grand total = 200 x 50 / 500 = 20. An answer of 25 assumes defects split evenly between shifts, ignoring the different volumes.
Expected count = row total x column total / grand total = 200 x 50 / 500 = 20. An answer of 25 assumes defects split evenly between shifts, ignoring the different volumes.
Question 98Define
A project charter states the relationship Y = f(X). In this expression, what does Y represent?
Correct answer: C. The project's output measure to improve
Y is the output or CTQ measure the project aims to improve, and the Xs are inputs and process factors that drive it. The input the team controls is an X, not Y.
Y is the output or CTQ measure the project aims to improve, and the Xs are inputs and process factors that drive it. The input the team controls is an X, not Y.
Question 99Improve
A team builds a stacked bar chart showing each operator's work content against the takt time line, and sees that one operator is far above takt while others are well below. What should the team do NEXT?
Correct answer: A. Rebalance tasks across operators
This operator balance chart, often called a yamazumi chart, shows imbalance, and tasks should be moved so each operator's load is at or below takt. Takt time is set by demand, so it cannot be changed to hide the imbalance.
This operator balance chart, often called a yamazumi chart, shows imbalance, and tasks should be moved so each operator's load is at or below takt. Takt time is set by demand, so it cannot be changed to hide the imbalance.
Question 100Control
After a project updates a work instruction, staff at one site are found still using a printed copy of the old version. What control would BEST prevent this?
Correct answer: D. Document control with version management
Document control ensures only the current approved version is available and old copies are withdrawn. Reminders and extra copies do not stop outdated documents from being used.
Document control ensures only the current approved version is available and old copies are withdrawn. Reminders and extra copies do not stop outdated documents from being used.
The real Lean Six Sigma Green Belt exam at a glance
100questions
3htime limit
MCQmultiple choice
75%+aim for here
Score 75% or more on two different sets before you book the real exam. Need a plan for your weak topics? See how we help you pass, or get a full refund.
More free Six Sigma & ITIL practice tests
Lean Six Sigma Black Belt Practice Test150 questions · DMAICITIL 4 Foundation Practice Test40 questions · ITIL 4
See all 36 free practice tests →
Original practice questions written by CertOcean in the style of the Lean Six Sigma Green Belt exam. They are not real exam questions. Lean Six Sigma Green Belt and Lean Six Sigma are trademarks of their owners. Real exam format: 100 questions · 3 hours · about 20 questions per DMAIC phase · common pass mark around 70%.