Free ASQ Certified Six Sigma Black Belt Sample Questions

Correct answer: A — True

• A. Both insufficient management support and insufficient dedicated resources such as time and budget are consistently identified as leading causes of failed Six Sigma deployments.
• B. This statement matches well-documented deployment failure analysis, so labeling it false would misrepresent two of the most frequently cited root causes.

Correct answer: A — Use Six Sigma's DMAIC and statistical analysis for the solder-joint failures, and use Lean tools to remove waste and cut changeover time on the assembly line

• A. This pairing matches the tool to the nature of each problem: DMAIC's statistical rigor is best suited to an unclear, variable-driven root cause, while Lean waste-removal tools directly address the flow and changeover issues on the assembly line.
• B. 5S is a housekeeping and standardization tool, not a method for isolating an unclear, multi-variable root cause, and hypothesis testing is unnecessary overhead for a changeover problem whose cause is already known.
• C. Applying a full statistical DMAIC cycle to a problem whose cause is already understood, such as the changeover delay, adds unnecessary time and cost when a direct Lean fix would work.
• D. Value stream mapping surfaces waste and flow issues but is not designed to isolate a statistically unclear, multi-variable root cause like the solder-joint failures, so it would not resolve that problem as a side effect.

Correct answer: B — False

• A. Six Sigma projects are meant to deliver value across stakeholders, particularly the customer; meeting an internal metric while measurably worsening customer outcomes does not represent genuine project success.
• B. This is correct: Six Sigma ties process performance to stakeholder value, especially customer impact, so a project that hits an internal target but degrades the customer experience has not truly succeeded.

Correct answer: A — Competitive benchmarking

• A. Comparing performance directly against named rivals in the same market on the same parameter is the defining feature of this benchmarking type. The bank is ranking itself against specific same-industry competitors rather than studying general practices.
• B. This would involve studying documented methods from any high-performing organization, including ones outside banking, rather than ranking against three specific same-market rivals. The scenario names direct competitors, which points elsewhere.
• C. This requires the organizations involved to jointly agree to share data and co-develop improvements, which is not described here since the bank is simply ranking itself using public rate sheets.
• D. This is aimed at crossing industry boundaries to pursue dramatic gains from an unrelated field, but the bank is comparing itself only to same-industry rivals, not an outside industry.

Correct answer: A — Average hold time is a leading indicator because it can be influenced now to change a future outcome, while churn rate is a lagging indicator because it reports a result that already occurred.

• A. Average hold time is actionable now and predicts a future result, which is the defining trait of a leading indicator, while churn rate reports what has already happened and cannot be changed retroactively, which is the defining trait of a lagging indicator. This is the correct classification for the two measures described.
• B. Being measured in real time does not make a metric lagging, and predicting call volume is not what churn rate does; churn rate reports a past outcome, so it is lagging, not leading. This option swaps the classifications incorrectly.
• C. Reporting cadence alone does not determine whether a measure is leading or lagging; the distinction depends on whether the measure predicts a future outcome or reports a past one. Churn rate reports a past outcome, so it cannot be classified as leading.
• D. Whether a measure ties to a specific balanced scorecard perspective is unrelated to the leading versus lagging distinction, and average hold time is actionable in advance, so it is not lagging. This reasoning applies the wrong criterion entirely.

Correct answer: A — Providing subject-matter guidance on Six Sigma methods and tools so the team applies them correctly during the project.

• A. Guiding the team on the correct application of Six Sigma tools and methods is the core purpose of the coach role, distinct from directing daily tasks or approving budgets.
• B. Assigning daily technical tasks describes the leader's role in directing project execution, not the coaching function of guiding correct tool and method use.
• C. Recording minutes and distributing action items is an administrative or scribe function, not the subject-matter guidance role that defines the coach.
• D. Approving the charter and allocating budget describes sponsor or Champion responsibilities, not the coach's role of guiding correct methodology.

Correct answer: A — Storming, where members test boundaries and compete over roles while working through disagreement

• A. Correct. Open disagreement over methods and competition for informal leadership, while the team stays committed to its charter, are hallmark behaviors of the storming stage of team development.
• B. Incorrect. Forming is characterized by members still getting acquainted and avoiding direct conflict, which contradicts the open disagreement and leadership competition described in the scenario.
• C. Incorrect. Norming is marked by acceptance of differing perspectives and smoother cooperation, which is inconsistent with the active competition for leadership described here.
• D. Incorrect. Performing teams operate with minimal role conflict and focus on shared goals, which does not match the described competition over informal leadership and tool selection.

Correct answer: A — Nominal group technique, where each member silently writes ideas independently before the group collects and ranks them.

• A. Nominal group technique has members generate ideas independently and anonymously before any group discussion, which prevents the senior engineer's presence from suppressing junior engineers' contributions. This directly solves the dominance problem described in the scenario.
• B. Open floor brainstorming relies on members calling out ideas in real time, which tends to favor the most vocal or senior person in the room and would likely reproduce the same dominance pattern. It does not create the independent generation step needed here.
• C. Multi-voting narrows a list that already exists rather than generating new ideas, so it does not address the problem of junior engineers failing to contribute ideas in the first place. It is a later-stage tool, not an idea-generation tool.
• D. The Delphi method is normally conducted anonymously and asynchronously across rounds rather than through live, real-time debate, and holding it as a live video debate would reintroduce the same dominance dynamic the technique is meant to avoid. Live debate contradicts the anonymous, independent structure of the method.

Correct answer: A — Conduct a skills gap analysis comparing each member's current competencies against the skills the project requires

• A. Identifying the gap between current and required skills is the first step of an effective training curriculum; it determines who needs what training before objectives or materials are set.
• B. A facilitation workshop addresses a soft skill that was not identified as a gap and does not resolve the missing statistical competencies the project actually needs.
• C. A blanket refresher ignores that only some members lack specific skills, wasting time for competent members and possibly under-training the two who need it most.
• D. Materials cannot be developed effectively before the gaps are identified and learning objectives are set, since the content would not target the actual deficiencies.

Correct answers: A, B, D — Adults engage more readily when training addresses a problem they need to solve right now; Adults' prior work experience should be used as a resource rather than dismissed as irrelevant; Adults generally prefer being told exactly why a topic matters before being asked to learn it

• A. Correct. Readiness to learn is tied to solving an immediate, real problem, which is a core andragogical principle.
• B. Correct. Using the learner's accumulated experience as a resource rather than ignoring it is one of Knowles' six principles of andragogy.
• C. Incorrect. Adult motivation is described as predominantly intrinsic; external disciplinary threats are not the primary driver in andragogical theory.
• D. Correct. The need-to-know principle holds that adults want to understand the reason for learning something before investing effort in it.
• E. Incorrect. Andragogy explicitly contrasts with pedagogy by reducing directive, teacher-led structure in favor of self-directed, autonomous learning for adults.

Correct answer: A — Multiply the relationship scores by the customer importance ratings to calculate weighted priorities for each technical requirement.

• A. Multiplying the relationship scores by the customer importance ratings produces a weighted priority for each technical requirement, which is exactly how QFD translates voice of the customer data into a ranked engineering focus list.
• B. Technical requirements often exist to satisfy customer needs that were not stated in exact wording, so discarding anything not literally mentioned would remove requirements the relationship matrix has already shown to matter.
• C. Spreading resources evenly ignores the very relationship and importance data the team just calculated, defeating the purpose of building the matrix to identify where effort will have the greatest customer impact.
• D. Substituting engineering difficulty estimates for customer importance ratings shifts the prioritization away from the voice of the customer, which is the input QFD is specifically designed to preserve through the analysis.

Correct answer: A — A simple Gantt chart does not show task dependencies unless linking arrows are added

• A. A basic Gantt chart plots task bars against time but does not natively encode which tasks depend on which others; dependency arrows or a linked network view must be added to show that relationship. This is exactly the gap the overlapping, unlinked bars reveal.
• B. A Gantt chart can display multiple concurrent tasks within the same time period without limitation; the scenario's issue is missing dependency information, not a display capacity limit.
• C. A basic Gantt chart does not automatically compute or recalculate a critical path; that capability typically requires additional scheduling logic or software beyond a simple bar chart.
• D. A Gantt chart does not require a RACI owner to be assigned before a task bar can be plotted; scheduling and responsibility assignment are separate, independent PM tools.

Correct answer: A — It states the measurable difference between current baseline performance and the improvement goal without naming a cause or a fix

• A. Quantifying the gap between baseline and goal while staying neutral on cause and solution is the defining feature of a properly scoped problem statement in the define phase.
• B. Naming a responsible department assigns blame before data has been collected, which is a common flaw reviewers are trained to catch during charter reviews.
• C. Recommending a specific fix skips the measure and analyze phases entirely, prejudging a solution the team has not yet validated with data.
• D. Listing planned measure and analyze tools describes project methodology rather than describing the performance gap the problem statement exists to capture.

Correct answer: A — An affinity diagram, sorting the ideas silently into related clusters and labeling each

• A. An affinity diagram is built by having the team silently group a large volume of unstructured ideas into natural clusters and then labeling each cluster, which is exactly the situation of many loosely worded ideas with no existing theory. It is the standard first step for organizing qualitative brainstorming output before deeper analysis begins.
• B. A tree diagram decomposes an already-defined objective or problem statement into progressively finer sub-causes or sub-tasks, which requires some structure to decompose. It is not designed to sort a large unstructured pile of raw, unrelated ideas into groups.
• C. An interrelationship digraph maps cause-and-effect arrows between a manageable set of already-identified factors to find key drivers. Applying it directly to over 100 raw, unsorted ideas is impractical before those ideas have been thematically grouped.
• D. A prioritization matrix ranks a defined set of options against weighted criteria using numerical scoring. It assumes the options are already discrete and organized, not a raw pile of overlapping brainstorming statements.

Correct answer: A — True

• A. This is correct. Takt time equals available production time divided by customer demand, and it sets the required production pace; it is a target rate, distinct from the process's actual measured cycle time.
• B. This is incorrect. The formula and interpretation described (available time divided by demand, used as the required pace) match the standard definition of takt time exactly, so the statement is true rather than false.

Correct answer: A — The measurement system consumes too much of the tolerance band and is unsuitable for this specification without improvement.

• A. This is correct because P/T ratios follow the same guidance bands as percent R&R, and a value of 0.35 exceeds the roughly 0.30 threshold where the measurement system is considered to consume too much of the tolerance and needs improvement.
• B. A 0.5 cutoff is far too loose; industry guidance treats values above roughly 0.30 as unacceptable, so this threshold misstates the standard.
• C. The P/T ratio is a variable-data metric derived from R&R standard deviation and tolerance width, and it applies directly to the gauge described here, not exclusively to attribute studies.
• D. P/T ratio compares measurement precision to the specification width; it is not a substitute for or limited to a stability study, which instead tracks bias over calendar time.

Correct answer: A — Each measurement can be linked through an unbroken chain of comparisons to a national or international reference standard.

• A. This is correct because traceability specifically refers to an unbroken chain of calibration comparisons connecting a working device back to a higher-level, ultimately national or international, reference standard.
• B. A unique identifier supports asset tracking and calibration records but does not by itself establish the chain of comparisons that defines traceability.
• C. Operator training relates to measurement system reproducibility and competence, not to the metrological chain linking a device to a reference standard.
• D. Storing results electronically supports data retention and audit trails, which is a documentation practice separate from the calibration chain that establishes traceability.

Correct answer: A — Nominal data labels categories with no inherent order, such as defect type, while ordinal data ranks categories, such as low, medium, or high satisfaction.

• A. This is correct because nominal scales simply name unordered categories, such as defect type, while ordinal scales rank categories in a meaningful order, such as low, medium, and high satisfaction.
• B. This is incorrect because nominal categories are unordered labels regardless of whether numeric codes are used, and equal-interval spacing is a property of interval or ratio scales, not nominal scales.
• C. This is incorrect because magnitude and frequency are quantitative measurements, not the defining property that separates nominal labeling from ordinal ranking.
• D. This is incorrect because a true zero point is a property of ratio scales, and neither nominal nor ordinal scales are defined by zero points or negative values.

Correct answer: A — The interquartile range, because it is based on the middle 50% of the data and is not influenced by extreme values in the tails.

• A. The interquartile range only uses the values between the 25th and 75th percentiles, so extreme outlier days sitting far in the tails do not affect it. This makes it the preferred measure of routine spread when a few unrepresentative extreme points are present.
• B. Standard deviation squares deviations from the mean, so it is highly sensitive to extreme values rather than resistant to them; equal weighting of every point means outliers pull it upward, not that it reflects their impact fairly. This is precisely why standard deviation is a poor choice when outliers should be excluded from the routine-spread picture.
• C. The range uses only the two most extreme points in the dataset, so a single outlier day can dominate the entire statistic. This makes it more distorted by outliers than almost any other spread measure, not less.
• D. Squaring deviations in the variance calculation amplifies the influence of large deviations rather than reducing it, since bigger gaps from the mean become proportionally larger after squaring. Outlier days would therefore inflate the variance even more than they inflate the standard deviation.

Correct answer: A — True

• A. This is true because mutually exclusive events with nonzero probabilities cannot occur together, meaning the joint probability is zero. Independence requires the joint probability to equal the product of the individual probabilities, which cannot be zero when both individual probabilities are nonzero, so the two properties are mutually incompatible.
• B. This is not correct because the mathematical definitions of independence and mutual exclusivity conflict whenever both events have nonzero probability. Treating them as compatible ignores that mutual exclusivity forces a zero joint probability, which independence cannot satisfy under those conditions.

Correct answer: A — 0.75

• A. Ppu = (110 - 101) / (3 x 4) = 0.75 and Ppl = (101 - 90) / (3 x 4) = 0.92; Ppk is the smaller of the two, so it equals 0.75, reflecting the tighter margin toward the upper specification limit using overall variation.
• B. This value is Ppl, the ratio on the lower-limit side, not the smaller value that defines Ppk. Reporting the larger of the two one-sided ratios overstates the true long-term performance.
• C. This value is roughly what would result from using the full specification width divided by six standard deviations (a Pp-style calculation) rather than the one-sided, mean-referenced Ppk calculation, and it ignores the off-center mean entirely.
• D. This value does not correspond to either one-sided ratio for the given mean, standard deviation, and specification limits, and likely results from a rounding or arithmetic error in the denominator.

Correct answer: A — A chi-square goodness-of-fit test against the expected proportions

• A. This is correct: the chi-square goodness-of-fit test compares observed category counts against an expected distribution, which is exactly the comparison being asked about across the four complaint categories.
• B. A t-test compares means of continuous measurements between groups, not categorical counts against an expected distribution shape.
• C. ANOVA compares group means, but complaint category counts are frequencies rather than a continuous outcome with group means to compare.
• D. Restricting the analysis to only two of the four categories would discard information from the other categories and would not test the full expected distribution.

Correct answer: A — The true relationship between predictor and response is curved, so a linear model is inadequate and a higher-order term or transformation is needed

• A. This is correct; a systematic curved pattern in a residual-versus-predictor plot signals that a straight-line model is missing curvature in the true relationship, calling for a quadratic term or transformation.
• B. Unequal variance typically shows up as a funnel-shaped spread rather than a systematic curved trend, so this option describes a different diagnostic pattern.
• C. Serial correlation is diagnosed by plotting residuals in observation or time order and looking for runs or cycles, not by plotting residuals against the predictor variable.
• D. Sample size alone does not produce a systematic curved pattern in residuals; a curved pattern reflects a model form problem rather than an insufficient number of observations.

Correct answer: A — Vendor dependency is supplier risk, the outdated server is cyber-security risk, and inconsistent labeling is operational risk

• A. A single, unbacked software vendor is supplier risk, an outdated server vulnerable to breach is cyber-security risk, and inconsistent day-to-day labeling steps reflect operational risk in the process itself.
• B. This mismatches all three: vendor dependency describes a supply-continuity issue, not day-to-day operations, and the outdated server describes a digital vulnerability, not a vendor issue.
• C. This mismatches all three: the vendor concern is about supply continuity rather than digital access, and the outdated server is a digital vulnerability rather than a routine process issue.
• D. Lumping distinct concerns into one enterprise-level category loses the specificity needed to assign the right owner and mitigation for each distinct risk type.

Correct answer: A — The technician's repeated walking to the printer is motion, and carting the specimen trays to another building is transportation

• A. Motion is unnecessary movement of people during their work, matching the technician's repeated walking, while transportation is unnecessary movement of materials, matching the specimen trays carted between buildings.
• B. This reverses the definitions; a person's own movement is classified as motion, not transportation, and moving materials is transportation, not motion.
• C. Motion applies to the technician's own movement, but the cart of specimen trays is a materials move, which falls under transportation instead.
• D. Transportation applies to the specimen trays being moved, but the technician's own repeated walking is motion rather than transportation.

Correct answer: A — A randomized block design that blocks on shift, removing shift-driven residue differences from the estimate of experimental error

• A. Blocking on shift is appropriate because shift is a known source of variability in residue buildup; grouping parts by shift and comparing cleaning solutions within each shift removes that variability from the error term.
• B. Ignoring shift would leave a known, meaningful source of variation folded into experimental error, reducing the ability to detect real differences between the cleaning solutions.
• C. There is no meaningful raw material lot variation in this scenario, so blocking on a second nuisance variable that does not vary adds unnecessary complexity without benefit.
• D. Raw material lot is not a real source of variation here, so crossing it into the design would waste runs studying an effect that does not exist.

Correct answer: A — Redesign the bracket or fixture with an asymmetric feature so it can physically only be inserted in the correct orientation

• A. This is correct because an asymmetric feature makes the incorrect orientation physically impossible to install, which is a contact-type mistake-proofing solution that prevents the defect at its source rather than catching it afterward.
• B. A visual aid still depends on the operator noticing and correctly interpreting it every time, so it reduces but does not eliminate the chance of the same orientation mistake happening again.
• C. A peer double-check adds inspection effort after the bracket has already been installed, catching the defect rather than preventing the incorrect orientation from being possible in the first place.
• D. Increasing audit frequency only improves the odds of catching an already-defective unit further downstream, and it does not change anything about the assembly step that allows the mistake to occur.

Correct answer: A — Whether a solution's projected returns justify its projected costs before committing resources.

• A. A cost-benefit analysis weighs a solution's projected costs against its projected returns, helping the team decide whether the solution is worth pursuing.
• B. Checking whether data follows a normal distribution is a statistical assumption check performed during Measure or Analyze, not the purpose of a cost-benefit analysis.
• C. Measurement system accuracy and repeatability are assessed through a measurement system analysis, which is a separate tool from cost-benefit analysis.
• D. Selecting an appropriate control chart is a Control-phase decision about ongoing monitoring, not something a cost-benefit analysis addresses.

Correct answer: A — Overall equipment effectiveness is the product of availability, performance, and quality, so it is roughly 75% here

• A. Multiplying availability (420/480 = 87.5%), performance (90%), and quality (95%) is correct because overall equipment effectiveness is defined as that three-factor product, yielding approximately 74.8%, which is the standard TPM metric for tracking whether the improved process stays in control.
• B. Using first-pass quality alone is incorrect because it ignores the availability and performance losses captured in the downtime and reduced cycle speed, both of which are part of the six big losses TPM tracks.
• C. Dividing planned minutes by run minutes is incorrect because it inverts the availability ratio and cannot exceed 100%, misrepresenting a loss as a gain and omitting performance and quality entirely.
• D. Requiring a full breakdown is incorrect because overall equipment effectiveness is calculated continuously from availability, performance, and quality data regardless of whether a complete stoppage occurs.

Correct answer: A — A p chart, because it tracks the fraction nonconforming and is designed to accommodate a sample size that varies from period to period.

• A. Correct — a p chart plots the proportion of nonconforming items and explicitly allows the subgroup sample size to vary between periods, which fits a call center where daily call volume changes.
• B. Incorrect — an np chart plots the count, not the proportion, of nonconforming items and requires a constant subgroup size across periods, which does not hold when daily call volume fluctuates.
• C. Incorrect — a c chart counts defects per inspection unit under a constant area of opportunity, which fits something like scratches per panel, not a binary compliant/non-compliant classification with a varying sample size.
• D. Incorrect — compliant/non-compliant is attribute data, not a continuous measurement, so averaging it on an X-bar/R chart is not statistically appropriate and discards the binomial nature of the data.

Correct answer: A — Log the root causes, countermeasures, and results in a shared repository, tagging which countermeasures apply to similar lines.

• A. Capturing root causes, countermeasures, and results in a searchable repository, tagged for lines with matching conditions, is exactly how lessons learned are reinforced and replicated across a business. This turns a single project's gains into an organizational asset instead of a one-off fix.
• B. A verbal summary at closeout has no persistence and depends on individual memory and initiative to spread; it does not create a durable, searchable record other teams can act on later.
• C. A financial figure alone tells other lines that a gain occurred but gives them no root cause or countermeasure detail to replicate the fix, so it does not support reinforcement or replication.
• D. Waiting for a recurrence before documenting anything is reactive, not proactive, and defeats the purpose of lessons learned, which is to prevent the same problem from reappearing elsewhere in the first place.

Correct answer: A — Percent Gage R&R and number of distinct categories are calculated relative to total process or tolerance variation, so a narrower process variation changes those metrics even without any change to the gauge

• A. This is correct because %GRR and number of distinct categories are both ratios against total process or tolerance variation, so improving process capability changes those metrics even when the gauge's absolute performance has not changed.
• B. This is incorrect because calibration drift is a physical property of the instrument over time and use, not something that automatically results from an improved process's variation shrinking.
• C. This is incorrect because there is no general regulatory requirement tying a new MSA certificate to re-centering a control chart; reanalysis is a measurement-capability decision, not a filing mandate.
• D. This is incorrect because operator turnover is not the reason measurement capability must be reevaluated; the statistical relationship between gauge variation and process variation is what drives the need to reanalyze.

Correct answer: A — Enter the Optimize phase and run statistically designed evaluations across multiple performance levels to tune the sensitive tolerances before validating the design.

• A. This is correct: unpredictable tolerance behavior under varying conditions is precisely what the Optimize phase is designed to address, using designed experiments to tune the sensitive parameters across performance levels before the design proceeds to validation.
• B. This is incorrect: skipping Optimize and moving straight to Validate ignores the DMADOV sequence and risks validating a design whose tolerances have not yet been stabilized against noise factors.
• C. This is incorrect: tolerance sensitivity to temperature is a design-optimization issue, not a sign that customer requirements were captured incorrectly, so returning to Measure does not address the actual problem.
• D. This is incorrect: an unresolved tolerance issue is a design-parameter problem to be handled in Optimize, not evidence that the project's scope or charter from the Define phase was wrong.

Correct answer: A — Relocate the seal behind an accessible service panel so technicians can replace it without removing the motor housing

• A. This is correct because Design for Maintainability focuses on reducing the effort and time required to service a component, and relocating the seal behind an accessible panel removes the need for the housing removal step entirely. That directly shortens repair time and lowers the labor cost driving lifecycle service cost.
• B. Simply changing the documented interval does not alter the physical design, so when replacement eventually is needed, the motor housing must still be fully removed, leaving the underlying serviceability problem in place.
• C. A longer-life material reduces how often the seal fails but leaves the housing-removal access problem unchanged for the times replacement is still required, so it only delays the maintainability issue rather than resolving the access design.
• D. Adding a redundant seal inside the same housing increases part count and cost without changing how the worn seal is accessed, so a technician would still need to remove the housing whenever the primary seal is eventually serviced.

Correct answer: A — The smaller-is-better ratio, because the team's goal is to drive the response toward zero variation rather than hit a nominal target or a maximum.

• A. This is correct because the smaller-is-better ratio is used when the objective is to minimize a response value as close to zero as possible, matching a stopping-distance variation goal with no fixed target.
• B. The nominal-is-best ratio applies when there is a specific target value the response should hit, but this scenario has no target value, only a goal of minimizing variation itself.
• C. The larger-is-better ratio applies when a higher response value is always more desirable, which does not match a goal of minimizing stopping-distance variation.
• D. The dynamic ratio applies when a proportional relationship between an input signal and output response is being characterized, not when the sole objective is minimizing variation around zero.