Quality Basics
Quality Basics is the foundation domain of the CQPA exam. It tests whether you can define quality precisely, explain the major quality philosophies, classify cost-of-quality elements, and connect customer requirements to the processes that satisfy them. Expect definition-based items and short scenarios that ask you to choose the best classification or principle.
Definitions and philosophies
Quality is the degree to which a set of inherent characteristics fulfills requirements (ISO 9000). Two consequences follow: quality is judged against requirements, not against luxury, and grade is not quality-a high-grade product that misses requirements is still poor quality. Juran's trilogy divides quality management into planning, control, and improvement. Deming emphasized PDSA, constancy of purpose, and reducing reliance on mass inspection. Crosby defined quality as conformance to requirements and promoted zero defects and prevention. Feigenbaum introduced total quality control and popularized cost of quality.
Cost of quality
Four categories: prevention (training, design reviews, supplier development), appraisal (inspection, test, calibration), internal failure (scrap, rework, downtime), and external failure (warranty, returns, recalls, complaint handling). Prevention plus appraisal is the cost of good quality; internal plus external failure is the cost of poor quality (COPQ). The strategic aim is to shift spending toward prevention, which typically lowers total cost.
Customers and standards
Voice of the customer (VOC) is translated into critical-to-quality (CTQ) characteristics. Customers can be internal (next process) or external. ISO 9001 is built on seven principles: customer focus, leadership, engagement of people, process approach, improvement, evidence-based decision making, and relationship management. Sector standards such as IATF 16949, AS9100, and ISO 13485 add industry-specific requirements.
Worked example
A plant spends $40,000 on inspection, $25,000 on scrap, $10,000 on warranty, and $15,000 on training. Prevention = $15,000; appraisal = $40,000; internal failure = $25,000; external failure = $10,000. Total cost of quality = $90,000. COPQ = $35,000, or 38.9% of total. The largest single category is appraisal, which suggests the process relies on detection rather than prevention.
Traps that cost marks
Labeling training or supplier development as failure cost. Calling inspection "prevention." Treating a high-grade product as high quality when it misses requirements. Forgetting that external failure includes lost future business, not just warranty claims.
Teams and Team Processes
Teams and Team Processes tests how a quality analyst launches, facilitates, and sustains teams. Items cover team types, roles, development stages, communication, conflict, decision making, and meeting discipline. The exam favors scenario questions: a team behaves a certain way, and you must name the stage or select the leader's best response.
Team types and roles
Common types are project or improvement teams (time-bound, specific deliverable), cross-functional teams (members from several functions), self-directed work teams (ongoing operational ownership), and virtual teams. A sponsor or champion removes barriers and secures resources; the leader or facilitator guides process; the recorder captures decisions and actions; the timekeeper protects the agenda. A charter states purpose, scope, deliverables, milestones, resources, and decision rights.
Stages of development
Tuckman's stages are forming (polite, dependent), storming (conflict over roles and ideas), norming (agreement on rules and methods), performing (focus on results), and adjourning (disbanding and lessons learned). Storming is normal and often productive; the leader's job is to surface disagreement, clarify roles, and enforce ground rules rather than suppress conflict.
Communication and meetings
Communication requires a sender, message, channel, receiver, and feedback; noise and filtering are barriers. Active listening means paraphrasing and confirming. Effective meetings have an agenda sent in advance, ground rules, minutes, and an action log with owner and due date.
Conflict and decisions
Functional conflict tests ideas; dysfunctional conflict attacks people. The Thomas-Kilmann modes are competing, collaborating, compromising, avoiding, and accommodating; collaborating is best when both the issue and the relationship matter. Consensus means everyone can support the decision, not that everyone's first choice wins; unanimity is stronger. Brainstorming separates idea generation from evaluation; nominal group technique and multivoting help prioritize.
Worked example
A newly chartered team misses its second milestone, members debate whether the leader should even exist, and two subgroups form. This is storming, not norming. The best response is to revisit the charter and ground rules, clarify roles, and facilitate a discussion of the disagreements-not to replace the leader or force a vote.
Traps that cost marks
Calling early conflict "norming." Treating all conflict as disloyalty. Equating consensus with unanimity. Criticizing ideas during brainstorming. Skipping the adjourning lesson-learned step.
Continuous Improvement
Continuous Improvement tests the methods a quality process analyst uses to make processes better: improvement cycles, Lean and Six Sigma concepts, and the seven basic quality tools. Questions often ask which tool fits a situation or which step comes next in a cycle.
Improvement cycles
PDCA (plan-do-check-act), also called PDSA, is the underlying cycle: plan the change and predict results, do it on a small scale, check results against the prediction, and act by standardizing or adjusting. DMAIC (define, measure, analyze, improve, control) is the Six Sigma project structure. Kaizen is continuous, incremental improvement by the people doing the work; a kaizen event is a focused, short-duration workshop.
Lean and Six Sigma
Lean attacks waste-transportation, inventory, motion, waiting, overproduction, overprocessing, and defects (TIMWOOD). Core Lean tools include 5S, value stream mapping, kanban, poka-yoke (error-proofing), SMED (quick changeover), and total productive maintenance. Six Sigma attacks variation and defects using DMAIC, statistics, and belt roles (champion, master black belt, black belt, green belt). Lean and Six Sigma are complementary: Lean improves flow, Six Sigma improves consistency.
The seven basic quality tools
Check sheet (structured data collection), flowchart (process steps), histogram (distribution shape), Pareto chart (prioritization by frequency or cost), cause-and-effect diagram (potential causes by category), scatter diagram (relationship between two variables), and control chart (stability over time). Stratification-separating data by source-is often added as an eighth tool.
Worked example
A defect log shows scratches 60, dents 25, paint defects 10, and other 5, for 100 total. Cumulative percentages are 60%, 85%, 95%, and 100%. A Pareto chart shows that scratches and dents account for 85% of defects, so the first improvement project should target those two categories rather than spreading effort evenly.
Traps that cost marks
Confusing Lean (waste and flow) with Six Sigma (variation and defects). Building a Pareto chart from categories with nearly equal counts. Using a control chart to prioritize causes rather than to detect instability. Skipping the Check step, so no one knows whether the change worked. Treating kaizen as a one-time project rather than an ongoing habit.
Corrective and Preventive Action
Corrective and Preventive Action (CAPA) tests whether you can separate fixing a problem from eliminating its cause-and whether you can run a disciplined, verifiable process. Scenario items ask you to classify an action or choose the next CAPA step.
Definitions
A nonconformity is a failure to meet a requirement. Correction fixes the detected nonconformity (rework, sort, retorque). Containment protects the customer while the investigation proceeds. Corrective action eliminates the cause of a detected nonconformity so it does not recur. Preventive action eliminates the cause of a potential nonconformity so it never occurs. ISO 9000 defines these terms, and exam items often hinge on the distinction.
The CAPA workflow
- Identify and document the problem with data (audit finding, complaint, nonconforming material report, SPC signal, risk analysis).
- Contain and correct immediately.
- Determine root cause using 5 Whys, cause-and-effect, fault tree, FMEA, or data analysis.
- Plan and implement the action, with owner and due date.
- Verify effectiveness with data-trend charts, audits, defect rates-not opinion.
- Standardize: update documents, train, and read across to similar processes.
Root cause quality
Distinguish the direct cause, contributing causes, and the root cause. "Operator error" is rarely a terminal root cause; ask why the system allowed the error, and consider error-proofing. Verify causes with evidence before acting. A CAPA that cannot be verified is not closed.
Worked example
A torque wrench is found out of calibration, and fasteners on shipped units may be loose. Containment: quarantine inventory and notify customers. Correction: retorque or rework affected units. Corrective action: revise the calibration schedule and add gauge control. Preventive action: apply the same control to all torque tools and add a poka-yoke that prevents use of an out-of-calibration tool. Verification: track calibration compliance and torque audit results for three months.
Traps that cost marks
Calling containment or rework "corrective action." Closing a CAPA without an effectiveness check. Ending 5 Whys at "carelessness." Confusing preventive action with a correction. Failing to read across to similar lines or products.
Data Management
Data Management tests how you collect, describe, and interpret data-and how much you trust the measurement behind it. Items cover data types, sampling, measurement systems, descriptive statistics, control charts, and capability.
Data types and scales
Variable (continuous) data come from measurement (diameter, temperature). Attribute (discrete) data come from counting or classification (pass/fail, number of defects). Scales are nominal (categories), ordinal (rank), interval (equal differences, arbitrary zero), and ratio (true zero). Operational definitions state exactly how each value is obtained so different people collect comparable data.
Sampling
A population is the full set; a sample is the subset measured. Random sampling gives every unit a known chance of selection. Systematic sampling takes every kth unit; stratified sampling samples within strata; cluster sampling selects groups. Convenience samples, nonresponse, and measurement bias distort results.
Measurement systems
Accuracy is closeness to the true value (low bias); precision is closeness of repeated readings (low variation). Repeatability is variation when one appraiser measures the same units repeatedly; reproducibility is variation between appraisers. Gauge R&R studies estimate both; AIAG guidance treats %GRR under 10% as acceptable, 10–30% as marginal, and over 30% as unacceptable. Calibration, resolution, stability, and linearity also matter.
Statistics, control, and capability
Descriptive statistics include mean, median, mode, range, variance, and standard deviation. Control charts separate common-cause from special-cause variation; variables charts include Xbar-R, Xbar-s, and I-MR; attribute charts include p, np, c, and u. Cp = (USL − LSL) / 6σ measures spread against tolerance; Cpk = min[(USL − mean), (mean − LSL)] / 3σ also accounts for centering.
Worked example
A process has mean 10.00, σ 0.05, USL 10.30, LSL 9.70. Cp = 0.60 / 0.30 = 2.00. Cpk = min(0.30, 0.30) / 0.15 = 2.00. The process is both capable and centered. If the mean drifted to 10.10, Cpk would fall to min(0.20, 0.40) / 0.15 = 1.33 while Cp stayed at 2.00-showing a centering problem.
Traps that cost marks
Treating specification limits as control limits. Calculating capability from an unstable process. Putting attribute data on an Xbar-R chart. Ignoring measurement variation before reacting to a small shift. Using a convenience sample and calling it random.
Process Management
Process Management tests the process view of work: how inputs become outputs, how processes are mapped, controlled, documented, and improved, and how risk is managed. Items ask you to read a process description and choose the best control, measure, or improvement.
The process approach
A process is a set of interrelated activities that converts inputs into outputs. SIPOC (suppliers, inputs, process, outputs, customers) defines boundaries before detailed mapping. A process owner is accountable for performance across functions. Value-added steps transform the product or service in a way the customer pays for; non-value-added steps are waste. Process cycle efficiency = value-added time / total lead time.
Control and standardization
Standard work defines the current best method; control plans list the characteristic, specification, measurement method, sample size and frequency, and reaction plan. KPIs should measure outcomes (defects, cycle time, cost) rather than mere activity. Document control covers approval, revision, distribution, and retention. Internal audits check conformance and effectiveness; findings feed CAPA.
Risk and FMEA
Risk-based thinking asks what could go wrong and how likely and severe it would be. FMEA assigns severity, occurrence, and detection ratings (typically 1–10) and multiplies them into a risk priority number (RPN). High-RPN items get action, and the FMEA is updated after changes. Mistake-proofing (poka-yoke) is preferred over warnings and training because it makes errors impossible or immediately visible.
Worked example
An order-to-cash process has 10 days total lead time, of which 2 days are value-added. Process cycle efficiency = 2/10 = 20%. Mapping the as-is process reveals 4 days of waiting for email-based credit approval. A rule-based approval and daily queue review could cut 3 days, raising efficiency to 2/7 = 28.6%. The measure confirms improvement in flow, not just activity.
Traps that cost marks
Mapping the intended ("should be") process instead of the actual as-is process. Measuring busy-ness rather than outcomes. Writing documents that no one controls or updates. Treating FMEA as a one-time exercise. Choosing training as the only control when error-proofing is feasible.