I85PREP
⌘ K
QETPP study materials
QETPP study resources

QETPP practice questions

One question at a time. Build your QETPP understanding.

First 50 of 200 questions. Each one lists the choices, the correct answer, and an explanation. The checker at the top of this page saves progress in this browser when it is available.

Continue in the practice tool above to load more questions, 50 at a time.

Question 1Unverified - official content outline not available for this exam code

In a structured test process, which test level is primarily responsible for confirming that an individual module behaves as its detailed design specifies before it is combined with other modules?

  1. Component testing Correct answer.
  2. Acceptance testing
  3. System testing
  4. Integration testing

Explanation

Component (unit) testing isolates the smallest testable piece of software, such as a module, program or class, and checks it against its detailed design, often using stubs and drivers in place of collaborators. Integration testing only begins once components are combined, so it addresses interfaces and data flow rather than single-module behaviour.

Question 2Unverified - official content outline not available for this exam code

Which defect attribute expresses how badly a failure affects the operation of the system, independent of how soon the team intends to fix it?

  1. Priority
  2. Detectability
  3. Severity Correct answer.
  4. Reproducibility

Explanation

Severity classifies technical impact, such as a crash, data loss or a cosmetic error, whereas priority records business and scheduling urgency. The two often diverge: a crash on a rarely used screen is high severity but may be scheduled with low priority.

Question 3Unverified - official content outline not available for this exam code

A review confirms that a delivered system satisfies the users' real operational needs rather than merely conforming to the written specification. Which activity does this describe?

  1. Validation Correct answer.
  2. Static analysis
  3. Verification
  4. Regression testing

Explanation

Validation asks whether the right product was built, checking fitness for intended use against genuine user needs. Verification asks whether the product was built right, that is, whether it conforms to the requirements that were specified.

Question 4Unverified - official content outline not available for this exam code

A function contains a single IF statement with no ELSE, and a test suite executes every line of that function. The IF condition has only ever been evaluated as true. Which coverage status is accurate?

  1. Decision coverage is 100% and statement coverage is incomplete
  2. Both statement and decision coverage reach 100%
  3. Statement coverage is 100%; decision coverage is not Correct answer.
  4. Neither coverage measure has been achieved

Explanation

Decision coverage requires every decision to take both a true and a false outcome at least once, so if the IF condition only evaluated true, the implicit false path was never exercised and decision coverage for it remains at 50%. Statement coverage only requires each executable line to run, which the suite already achieves.

Question 5Unverified - official content outline not available for this exam code

A team adds a new discount rule to an existing billing module. Which activity is specifically intended to reveal whether previously working billing functions now behave differently?

  1. Usability testing
  2. Regression testing Correct answer.
  3. Alpha testing
  4. Static code review of the new rule

Explanation

Regression testing re-executes tests for functionality that already worked, in order to detect side effects introduced by a change such as a new discount rule. Reviewing only the new rule examines the change itself and cannot show how surrounding functions now behave.

Question 6Unverified - official content outline not available for this exam code

In risk-based testing, how is the risk level of a requirement normally determined?

  1. From the likelihood of failure multiplied by its impact Correct answer.
  2. From the severity ratings that developers assign during code review
  3. From the priority the product owner gives to the related user story
  4. From the number of test cases already automated for it

Explanation

Risk in testing is conventionally the product of likelihood, the probability that a defect will produce a failure, and impact, the business or user consequence if it does. Testing effort is then allocated proportionally to that combined level, so neither factor alone defines risk.

Question 7Unverified - official content outline not available for this exam code

During a nightly run, one automated check fails roughly one night in five with a different error each time, while the feature under test has not changed. What is the most defensible first action?

  1. Increase the retry count until the check passes consistently
  2. Raise a high-severity defect against the feature being tested immediately
  3. Delete the check and rely on manual testing of the feature
  4. Investigate the check for non-deterministic dependencies first Correct answer.

Explanation

An intermittent failure with varying errors and no corresponding product change points to non-determinism on the test side or in the environment, such as timing assumptions, shared test data or execution order, so the cause must be isolated before any result is trusted. Retrying or deleting the check hides the signal, and raising a defect is premature until the failure is shown to originate in the product.

Question 8Unverified - official content outline not available for this exam code

A release reports 100% statement coverage and full traceability to every requirement, yet customers still report serious defects. Which explanation is most defensible?

  1. Coverage and traceability measure test volume only, so adding more test cases would prevent the defects
  2. Statement coverage must have been miscalculated, since full coverage rules out defects in released code
  3. Traceability guarantees correctness, so the defects must relate to requirements added after release
  4. Execution was measured, but correct expected results and complete requirements were not Correct answer.

Explanation

Coverage metrics record which code was executed, and traceability records that tests exist for each requirement; neither shows whether the expected results in those tests were correct, nor whether the requirements themselves were complete and accurate. Missing requirements and weak oracles are therefore fully consistent with high reported coverage.

Question 9Unverified - official content outline not available for this exam code

Which characteristic best distinguishes exploratory testing from scripted test execution?

  1. Learning, design and execution overlap, with lightweight records Correct answer.
  2. Testing is performed without charters, session notes or any recording of results
  3. Tests are designed in advance and executed strictly in the order they were written
  4. Only automated checks are used, so that every result can be reproduced exactly

Explanation

Exploratory testing deliberately overlaps learning about the product, designing tests and executing them, with the tester's judgement steering each next step and documentation kept lightweight through charters and session notes. It is not unplanned testing, because charters, time boxes and recorded observations keep the sessions accountable.

Question 10Unverified - official content outline not available for this exam code

Which item is the most appropriate exit criterion for a system test phase?

  1. Open defects are below the agreed severity threshold Correct answer.
  2. The test plan has been reviewed and approved by the project manager
  3. The test environment has been built and smoke tests have passed
  4. Test cases have been designed and traced to the requirements baseline

Explanation

Exit criteria state what must be true before a test phase can be declared complete, and for system test this is normally expressed as defect status thresholds and agreed coverage targets being met. The remaining options describe planning milestones or entry criteria that are satisfied before execution even starts.

Question 11Unverified - official content outline not available for this exam code

Module A calls a payment service that has not been built yet. To run tests on module A now, the team writes a small component that accepts A's calls and returns fixed, pre-agreed answers. In test-double terminology, what is this component?

  1. A harness that launches the whole build automatically
  2. A driver that invokes the module under test
  3. A test oracle that decides whether each result is correct
  4. A stub that stands in for the module being called Correct answer.

Explanation

A stub is a test double that replaces a component called by the module under test and returns pre-programmed responses, so the caller can be exercised before the real component exists. Drivers play the opposite role: they call the module under test when that module is the lowest item in the integration hierarchy. An oracle supplies expected results and a harness is the automation scaffolding, so neither supplies canned answers in place of the missing callee.

Question 12Unverified - official content outline not available for this exam code

An input field accepts whole-number ages from 18 to 65 inclusive. A tester applies two-value boundary value analysis to this single valid range. Which set of values belongs in the test set?

  1. 18, 40 and 65, the endpoints plus a mid value
  2. 17 and 66 only, the invalid neighbours
  3. 17, 18, 65 and 66 Correct answer.
  4. 18 and 65 only, the two valid endpoints

Explanation

Two-value boundary value analysis tests each boundary together with its nearest neighbour outside the range, so the set is 17, 18, 65 and 66. Testing only 18 and 65 would leave the just-below and just-above values untried, which is exactly where off-by-one faults hide. Interior values add little because the partition is assumed to be processed identically throughout.

Question 13Unverified - official content outline not available for this exam code

Defect data from a long project shows that four of sixty modules account for most of the defects found, while the remaining modules contain very few. Which testing principle does this pattern illustrate?

  1. The pesticide paradox, since repeated tests stop finding new defects
  2. Defect clustering, since a few modules hold most defects Correct answer.
  3. Exhaustive testing is impossible for any non-trivial system
  4. Early testing, since defects are cheaper to fix the sooner they are found

Explanation

Defect data typically shows that a small number of components contain a disproportionate share of the defects, a pattern known as defect clustering. It is used to focus test effort, to re-test the riskiest modules hardest and to justify extra scrutiny of modules with a poor history. The other options are also testing principles, but none of them describes this uneven distribution of defects across modules.

Question 14Unverified - official content outline not available for this exam code

A shop gives a 10% discount only when the customer is a registered member and the order value exceeds 200 EUR; in every other situation no discount is applied. A tester wants a systematic set of cases for the two conditions and the resulting action. Which technique fits best?

  1. Boundary value analysis
  2. Decision table testing Correct answer.
  3. Equivalence partitioning
  4. State transition testing

Explanation

Decision table testing is the technique for combinations of conditions and the actions they trigger; two conditions with two values each yield four rule columns that become test cases. Equivalence partitioning and boundary value analysis work on the data ranges of a single input rather than on combinations of conditions. State transition testing models sequences of events over time, which is not what the discount rule describes.

Question 15Unverified - official content outline not available for this exam code

An ATM session is modelled with states IDLE, CARD_INSERTED, PIN_VERIFIED and DISPENSING, and the model forbids a withdrawal request while in CARD_INSERTED. A test enters CARD_INSERTED, requests a withdrawal and checks that the request is refused. Which coverage element does this test provide?

  1. Invalid transition coverage, since a forbidden transition is tested Correct answer.
  2. Valid transition coverage, because every allowed transition is executed
  3. State coverage, because every state is entered at least once
  4. Decision coverage, because each branch of the PIN check is taken

Explanation

Requesting a withdrawal before the PIN is verified is a transition the model forbids, so attempting it and confirming that the system rejects it is invalid transition coverage. Valid transition coverage exercises only the transitions the model permits. State coverage only requires each state to be reached, while decision coverage concerns branches in code, not the legality of a state change.

Question 16Unverified - official content outline not available for this exam code

A tool inspects the source without running it and reports a variable that is assigned a value but never read afterwards. What is the most accurate description of this result?

  1. Such variables are only detectable with a debugger during execution
  2. Execution produced it, so it counts as a dynamic defect
  3. No test failed, so it can be dismissed without review
  4. Static analysis produced it, and triage must confirm real dead code Correct answer.

Explanation

Static analysis inspects source code or models without executing the program, so a variable that is written but never read can be flagged before any test runs. The finding is a candidate defect rather than a proven one, because generated code, reflection or debugging hooks can make apparent dead code intentional; triage decides. Only dynamic techniques such as debugging or runtime tracing need an execution.

Question 17Unverified - official content outline not available for this exam code

A tester logs a defect that reads only: "Login is broken. Fix it." Which missing element most directly prevents a developer from reproducing the failure?

  1. The tester's name and the date and time on which the run was made
  2. The identifier of the requirement that the test case traced to
  3. The steps, data and environment used when the failure appeared Correct answer.
  4. The severity and priority agreed with the business owner

Explanation

A defect report is only actionable if the developer can reproduce the failure, which requires the steps, the input data and the environment or build in which it occurred. Name and date, severity and priority, and trace links support management and analysis but do not let anyone reproduce the failure. Without reproduction information the report is likely to be returned or closed as not reproducible.

Question 18Unverified - official content outline not available for this exam code

A nearly finished product is installed at customer sites and used by real operators in their own working conditions, with no developers present and no control over the environment. How is this test level best described?

  1. Interface testing, checking the exchange of data between two systems
  2. Beta testing, run at customer sites with real users and no developer control Correct answer.
  3. Alpha testing, performed at the developer's own site under controlled observation
  4. Smoke testing, a quick check that the main functions still work

Explanation

Beta testing places a nearly finished product with real users in their own environment, where the development team cannot control conditions and feedback comes from actual use. Alpha testing also involves users but takes place at the developer's site under controlled observation. Smoke and interface testing are checks of build stability and of data exchange between components, not user trials at customer sites.

Question 19Unverified - official content outline not available for this exam code

A control flow graph for one function has 7 nodes and 9 edges. Applying McCabe's formula V(G) = E − N + 2, how large a basis set of independent paths is needed?

  1. 16 basis paths
  2. 2 basis paths
  3. 3 basis paths
  4. 4 basis paths Correct answer.

Explanation

McCabe's cyclomatic complexity is V(G) = E − N + 2, so 9 edges and 7 nodes give 4. That value is the number of independent paths a basis set must contain so that every decision outcome is exercised at least once. It is not the total number of paths through the graph, which grows combinatorially with the number of decisions.

Question 20Unverified - official content outline not available for this exam code

A configuration screen has four parameters with three values each, giving 81 full combinations. The team decides to use an all-pairs (pairwise) design instead. Which statement correctly describes what the resulting suite guarantees?

  1. It guarantees detection of every defect caused by any interaction between the parameters.
  2. It guarantees every pair of values of any two parameters is covered, but not every triple. Correct answer.
  3. It guarantees that every combination of values of all four parameters is executed at least once.
  4. It guarantees that every individual value of each parameter is executed at least once, but nothing more.

Explanation

All-pairs, or pairwise, testing generates a set of cases in which every combination of values of any two parameters appears at least once. For four three-valued parameters that is far fewer cases than the 81 of the full cartesian product. It does not cover every triple or higher-order interaction, so a defect that appears only when three specific values coincide can still escape.

Question 21Unverified - official content outline not available for this exam code

A booking field accepts a passenger count as a whole number from 1 to 9. A tester divides the input into the classes 'less than 1', '1 to 9', and 'greater than 9'. Which test design technique is the tester applying?

  1. Boundary value analysis
  2. State transition testing
  3. Equivalence partitioning Correct answer.
  4. Decision table testing

Explanation

Equivalence partitioning divides the input domain into classes whose members are expected to be handled the same way, then selects representative values from each class. The three ranges described are exactly such classes, so one value from each is enough for this technique. Boundary value analysis would instead concentrate on values at 0, 1, 9, and 10.

Question 22Unverified - official content outline not available for this exam code

In a formal review, which role is responsible for keeping the review focused, ensuring that the review process is followed, and mediating disagreements between participants?

  1. The author
  2. The line manager
  3. The scribe
  4. The moderator Correct answer.

Explanation

The moderator, also called the facilitator, plans and leads the formal review, keeps participants on the review agenda, and helps resolve disagreements. The author explains the work product but does not control the review process. The scribe records findings, and the line manager usually has no operational role in the review meeting.

Question 23Unverified - official content outline not available for this exam code

A test team needs a stable, approved set of testware items that will serve as the basis for a formal test cycle. In configuration management, what is this approved set called?

  1. A baseline Correct answer.
  2. A build
  3. A change request
  4. A branch

Explanation

A baseline is an approved and relatively stable set of configuration items that is placed under formal change control and used as a reference for later work. Changes to a baseline normally require a change request and approval. A branch is a line of version development, a change request asks for a modification, and a build is an executable deliverable assembled from configuration items.

Question 24Unverified - official content outline not available for this exam code

A defect report claims that a screen shows a warning when it should not. The developer checks the approved specification and finds that the warning is required. Which defect status best reflects the developer's conclusion that the report is invalid?

  1. Reopened
  2. Rejected Correct answer.
  3. Deferred
  4. Duplicate

Explanation

A defect is rejected when the reviewer determines that the reported behavior is not a defect, for example because it matches the approved specification. Deferred means the defect is accepted but postponed, duplicate means it repeats an existing report, and reopened means a previously closed defect still occurs.

Question 25Unverified - official content outline not available for this exam code

A test manager estimates a regression cycle using three-point estimation. The optimistic estimate is 4 days, the most likely is 6 days, and the pessimistic is 14 days. Using the weighted formula (optimistic + 4 × most likely + pessimistic) / 6, what is the expected duration?

  1. 12 days
  2. 7 days Correct answer.
  3. 6 days
  4. 8 days

Explanation

Substituting the values gives (4 + 4 × 6 + 14) / 6 = (4 + 24 + 14) / 6 = 42 / 6 = 7 days. The weighted formula gives more influence to the most likely estimate while still accounting for uncertainty. The other values do not result from the stated formula.

Question 26Unverified - official content outline not available for this exam code

A team must verify a new tax calculation engine that applies many interacting rules. No independent expected-value table exists yet, and manually calculating each case is slow. Which approach provides the most practical test oracle?

  1. Re-run the same test cases until the outputs look plausible
  2. Compare results with an independently built spreadsheet model of the rules Correct answer.
  3. Ask the developer whether the output is correct for each execution
  4. Record the outputs of the first run and treat them as the expected results

Explanation

An independent spreadsheet model built from the tax rules can produce expected results without relying on the system under test, making it a practical oracle. The developer is not independent and asking for confirmation on every execution does not scale. A first run is only a valid oracle if its outputs are already known to be correct.

Question 27Unverified - official content outline not available for this exam code

A team wants to determine the maximum number of concurrent users the system can serve before response time degrades beyond acceptable limits. Which type of performance test is most directly aimed at finding that breaking point?

  1. Stress testing Correct answer.
  2. Load testing
  3. Smoke testing
  4. Soak testing

Explanation

Stress testing drives the system beyond expected load to find the point where performance becomes unacceptable or failures occur. Load testing normally checks behavior at expected load levels, and soak testing checks sustained load over a long period. Smoke testing only checks basic viability.

Question 28Unverified - official content outline not available for this exam code

During system testing, the team finds 60 defects. After release, users report 15 additional defects that were not found in testing. Using defect detection percentage (DDP) with defects found in testing divided by all defects found in testing and after release, what is the DDP?

  1. 75%
  2. 25%
  3. 20%
  4. 80% Correct answer.

Explanation

DDP = defects found in testing / (defects found in testing + defects found after release) × 100%. Here 60 / (60 + 15) = 60 / 75 = 0.80, or 80%. The other values misplace the numerator or denominator in the calculation.

Question 29Unverified - official content outline not available for this exam code

A team considers a capture/replay tool for regression testing of a web UI that changes layout and element identifiers every few sprints. Which risk is the most significant reason to question this tool choice?

  1. The tool will prevent testers from recording any network traffic
  2. Scripts will need frequent maintenance as UI object identifiers change Correct answer.
  3. The tool cannot be used for unattended test execution across browsers
  4. The tool will remove the need for manual exploratory testing

Explanation

Capture/replay tools record UI actions tied to object identifiers; when the UI changes, recorded scripts break and require maintenance. This makes them costly for frequently changing interfaces. They can typically store many scripts and run unattended, and they do not replace exploratory testing.

Question 30Unverified - official content outline not available for this exam code

A test environment is refreshed with a copy of production data that contains customer names, addresses, and card numbers. Which practice best protects personal data while keeping the tests realistic?

  1. Copy the full production database to each tester's personal laptop for local use
  2. Use the production data unchanged because test environments are not production
  3. Replace sensitive values with realistic but fictitious data before loading the test environment Correct answer.
  4. Restrict access to the production data to the test manager and the lead developer

Explanation

Data masking or anonymization replaces personal identifiers with realistic fictitious values so test data remains representative without exposing real people's data. Copying production data to laptops, limiting access to one person, or using raw production data in test all leave personal data exposed.

Question 31Unverified - official content outline not available for this exam code

A QETPP candidate reviews a requirements specification for consistency and checks that each design document can be traced to an approved requirement before any executable code exists. Which testing activity is being performed?

  1. Debugging
  2. Regression testing
  3. Validation
  4. Verification Correct answer.

Explanation

Verification confirms that work products meet specified requirements and are internally consistent, and it can be performed without executing code. Validation would instead check fitness for intended use against user needs. Regression testing and debugging are different activities with different objectives.

Question 32Unverified - official content outline not available for this exam code

Before running a new interest-calculation test, a tester obtains the approved formula, worked examples, and expected monthly payment table from the business analyst. In testing terminology, what does this material provide for judging actual results?

  1. Test charter
  2. Test procedure
  3. Test harness
  4. Test oracle Correct answer.

Explanation

A test oracle is the source or mechanism used to determine the expected result and judge whether an actual result is correct. The approved formula and worked payment table supply those expected values. A test procedure describes steps, a harness supports execution, and a charter guides exploratory testing.

Question 33Unverified - official content outline not available for this exam code

A project team tests two completed components together through their published interfaces, then adds a third component and repeats the checks. Which test level does this activity primarily represent?

  1. Integration testing Correct answer.
  2. Acceptance testing
  3. System testing
  4. Unit testing

Explanation

Integration testing focuses on interactions and interfaces between components or systems as they are combined. Adding components incrementally and checking their published interfaces is characteristic of integration testing. Unit, system, and acceptance testing address different scopes.

Question 34Unverified - official content outline not available for this exam code

A regulated project records each requirement identifier and links every test case to one or more requirements in the test management tool. Which benefit does this practice provide most directly?

  1. It supports coverage checks and change impact analysis. Correct answer.
  2. It replaces the need for regression testing after a change.
  3. It guarantees that every test case has a unique priority.
  4. It proves that linked requirements contain no defects.

Explanation

Requirements-to-test traceability shows which requirements are covered by tests and which have no linked tests. When a requirement changes, the links help identify the test cases that must be reviewed or rerun. Traceability does not prove absence of defects, replace regression testing, or assign priorities.

Question 35Unverified - official content outline not available for this exam code

A test manager has limited time before release. Two features remain: Feature X has a high likelihood of failure and high business impact; Feature Y has a low likelihood of failure but medium business impact. Following risk-based testing, which feature should receive deeper test effort first?

  1. Feature Y, because medium impact always outranks high impact.
  2. Feature X, because its risk exposure is higher. Correct answer.
  3. Neither feature, because limited time removes the need for prioritization.
  4. Either feature, because risk-based testing ignores likelihood.

Explanation

Risk-based testing prioritizes test effort according to risk exposure, commonly combining likelihood and impact. Feature X is high on both dimensions, so it has the greater risk exposure and should receive deeper testing first. Limited time makes prioritization more important, not less important.

Question 36Unverified - official content outline not available for this exam code

A tester is given a 90-minute time box, a written charter to investigate coupon-code abuse, and a debrief with the test lead afterward. Which test approach is being used?

  1. Formal inspection of the coupon-code source code
  2. Checklist-based testing with no exploratory element
  3. Regression testing driven by a fixed test suite
  4. Session-based exploratory testing Correct answer.

Explanation

Session-based exploratory testing structures exploratory work around a charter, a time box, notes or recordings, and a debrief. The tester simultaneously learns about the product, designs tests, and executes them within that session. The other options describe different review or test approaches.

Question 37Unverified - official content outline not available for this exam code

A defect makes the application crash whenever a user opens the 'Archived Orders' report. Only three internal auditors use that report, and the next customer release is due tomorrow. How should severity and priority most likely be classified?

  1. Low severity and high priority
  2. High severity and high priority
  3. High severity and low priority Correct answer.
  4. Low severity and low priority

Explanation

Severity reflects the technical impact of the defect, and a crash in a report is high severity because the function becomes unavailable. Priority reflects business urgency; few affected users and an imminent customer release can make it low priority relative to other work. Severity and priority are related but distinct attributes.

Question 38Unverified - official content outline not available for this exam code

After a requirement changes, a tester updates several test cases. The team must later reproduce the exact testware used in an earlier test cycle and compare it with the current version. Which configuration management practice makes this possible?

  1. Version control with baselines and change history Correct answer.
  2. Recording only pass/fail status from the earlier cycle
  3. Deletion of all old test cases after each cycle
  4. Email distribution of changed test cases without version labels

Explanation

Configuration management keeps testware under version control, records changes, and establishes baselines so earlier configurations can be identified and reproduced. This allows an earlier test cycle to be compared with the current version. Pass/fail records and unlabeled copies do not preserve test case content reliably.

Question 39Unverified - official content outline not available for this exam code

Consider the decision `if (age >= 18 && hasConsent)` with an else branch. A test set contains two cases: (age=20, hasConsent=true) and (age=20, hasConsent=false). Which coverage statement is accurate for this decision?

  1. Condition coverage, but not decision coverage.
  2. Both decision coverage and condition coverage.
  3. Decision coverage, but not condition coverage. Correct answer.
  4. Neither decision coverage nor condition coverage.

Explanation

The first case takes the true branch and the second takes the false branch, so decision or branch coverage is achieved. Condition coverage requires each basic condition to take both true and false outcomes; `age >= 18` is true in both cases, so condition coverage is not achieved.

Question 40Unverified - official content outline not available for this exam code

A layered system is integrated top-down, with stubs replacing lower layers during early cycles. Which statement describes the main trade-off of this strategy?

  1. No stubs or drivers are needed, so integration effort is always lower.
  2. Upper-layer logic is tested early, while real lower-layer behavior is delayed. Correct answer.
  3. All components are integrated at once, which removes the need for regression testing.
  4. Lower-layer interface defects are found earlier because stubs run real lower-layer code.

Explanation

Top-down integration allows high-level control logic and workflows to be exercised early by using stubs for lower layers. The trade-off is that real lower-layer modules and their interfaces are tested later, so some defects may be discovered later. Stubs simulate behavior and cannot reveal defects in the real lower layers.

Question 41Unverified - official content outline not available for this exam code

During a QETPP-aligned review, a team compares a payment module's observed behavior with the approved business requirements to confirm that the module supports the intended user need. Which concept does this activity primarily demonstrate?

  1. Debugging
  2. Validation Correct answer.
  3. Static analysis
  4. Verification

Explanation

Validation asks whether the product meets its intended use and stakeholder needs, whereas verification asks whether a work product conforms to its specified requirements. Comparing delivered behavior with approved business requirements to confirm the right product was built is therefore validation, not a code-level inspection or defect-location activity.

Question 42Unverified - official content outline not available for this exam code

A ticketing feature has a fixed set of statuses and permitted events, and testers must exercise both allowed transitions and attempts to trigger forbidden ones. Which black-box technique directly models this aspect of the feature?

  1. Decision table testing
  2. State transition testing Correct answer.
  3. Error guessing
  4. Equivalence partitioning

Explanation

State transition testing models a system as states, events, transitions, and guards, so it directly supports valid-transition and invalid-transition cases. Equivalence partitioning and decision tables handle data classes or condition combinations rather than sequences of state changes.

Question 43Unverified - official content outline not available for this exam code

A component under test depends on a tax service that is not yet available. The team needs a test double that supplies fixed indirect inputs to the component and does not verify interactions. Which type of test double fits this need?

  1. Driver
  2. Fake
  3. Mock
  4. Stub Correct answer.

Explanation

A stub provides canned responses or fixed indirect inputs to the component under test and is not used to verify interaction behavior. A driver calls the component from below, a mock verifies interactions, and a fake is a working lightweight implementation rather than a fixed-response substitute.

Question 44Unverified - official content outline not available for this exam code

A quantity field accepts whole numbers from 1 through 50 inclusive. Using two-value boundary value analysis for the valid range, which set of quantity values is most appropriate?

  1. {0, 1, 50, 51} Correct answer.
  2. {1, 2, 49, 50}
  3. {1, 50}
  4. {0, 51}

Explanation

Two-value boundary value analysis tests each boundary and the nearest value on the other side of that boundary. For the valid range 1 to 50, the relevant values are 0 and 1 at the lower boundary and 50 and 51 at the upper boundary.

Question 45Unverified - official content outline not available for this exam code

A loan rule has three independent binary conditions and one approval action. Before any simplification, how many rule columns does a complete decision table contain for these conditions?

  1. 6
  2. 3
  3. 8 Correct answer.
  4. 9

Explanation

A complete decision table with n independent binary conditions has 2^n rule columns before simplification or merging. With three conditions, that is 2^3 = 8 columns.

Question 46Unverified - official content outline not available for this exam code

In a QETPP post-release analysis, defects are grouped by the module in which they were introduced rather than by the module in which they were detected. Which analysis does this grouping most directly support?

  1. Mean time to repair
  2. Defect detection percentage
  3. Defect clustering Correct answer.
  4. Statement coverage

Explanation

Defect clustering examines where defects concentrate, often by origin module or component, to focus testing and root-cause work. Detection percentage compares defects found in testing with total defects, coverage measures exercised code, and MTTR measures repair time.

Question 47Unverified - official content outline not available for this exam code

A regression suite contains 5,000 UI tests, and only a small fraction fail historically. The team wants fast commit feedback while still running broader checks periodically. Which automation approach best follows risk-based continuous-testing practice?

  1. Run all 5,000 UI tests on every commit
  2. Replace the automated suite with manual regression testing
  3. Run a risk-based smoke subset per commit and schedule full regression Correct answer.
  4. Disable every historically failing test so that the pipeline stays green

Explanation

Risk-based continuous testing uses a fast, stable subset for immediate feedback and defers broader regression to a scheduled run. Running everything on each commit often creates excessive feedback delay, and disabling failing tests hides risk rather than managing it.

Question 48Unverified - official content outline not available for this exam code

A static analysis tool flags a possible null dereference in a code path that current production configuration does not reach. What is the most appropriate next action in a QETPP review?

  1. Close the finding without assessing future reachability
  2. Log it immediately as a critical production defect
  3. Rewrite the whole module to remove null handling
  4. Triage the finding and record a justified disposition Correct answer.

Explanation

Static analysis findings require triage because tools can report issues in unreachable or defensive code and can produce false positives. The team should confirm reachability, assess impact, and record whether the finding is fixed, accepted, or rejected with rationale.

Question 49Unverified - official content outline not available for this exam code

A mutation test changes `amount > 0` to `amount >= 0`. The test suite passes on both the original and the mutated code. What does this result most directly indicate about the suite?

  1. It does not distinguish the boundary change, so the mutant survives Correct answer.
  2. The mutant is equivalent because both expressions always behave the same
  3. The original code must contain a defect at amount = 0
  4. It has achieved complete statement coverage

Explanation

A surviving mutant means the test suite produced the same result on original and mutated code, so it did not detect the change. Here the expressions differ at amount = 0, so a test at that boundary could kill the mutant; survival indicates a gap in boundary sensitivity.

Question 50Unverified - official content outline not available for this exam code

For a decision with several independent conditions, which statement accurately distinguishes modified condition/decision coverage (MC/DC) from exhaustive multiple condition coverage?

  1. Multiple condition coverage requires each condition to affect the outcome independently
  2. MC/DC shows each condition independently affects the outcome without all combinations Correct answer.
  3. Multiple condition coverage is weaker than decision coverage
  4. MC/DC requires every possible 2^n combination of conditions to be executed

Explanation

MC/DC requires evidence that each condition independently affects the decision outcome while keeping the number of test cases practical. Exhaustive multiple condition coverage tests all combinations of condition outcomes, which grows as 2^n and is stricter in combination count.