Explanation / Rationale

a) Is not correct. Analyzing a defect is part of debugging, not testing
b) Is correct. Creating test data is a test implementation task
c) Is not correct. While a tester may need to identify a test item’s version for results reporting purposes, assigning a test item’s version is part of configuration management
d) Is not correct. Writing a user story is not a testing activity and should be done by the product owner

Explanation / Rationale

a) Is not correct. While this test does match the description of an integration test, it is a non-functional test
b) Is correct. This test matches the description of an integration test and it is a non-functional test
c) Is not correct. This test does not match the description of a component test and it is not a functional test
d) Is not correct. While this test is a non-functional test, it does not match the description of a component test

Explanation / Rationale

a) Is correct. The management decides about performing the review
b) Is not correct. The moderator, not the review leader should ensure the effective running of review meetings
c) Is not correct. The author fixes the work product under review
d) Is not correct. The manager monitors ongoing cost-effectiveness

Explanation / Rationale

There are three equivalence partitions, with the boundaries as shown:

• Invalid too low (0.4 and below)
• Valid (0.5 to 25.0)
• Invalid too high (25.1 and above)

Thus:
a) Is not correct. None of those four boundary values are included in this set of tests. These tests do cover the equivalence partitions
b) Is not correct. All these four boundary values are included in this set of tests, but two additional values are included, one for each boundary. These are the values associated with three-point boundary value analysis
c) Is correct. Each of those four two-point boundary values are included in this set of tests
d) Is not correct. These four values are all included in the valid partition

Explanation / Rationale

a) Is not correct. A path through source code is one potential route through the code from the entry point to the exit point that could exercise a range of decision outcomes. Two different paths may exercise all but one of the same decision outcomes, and by just changing a single decision outcome a new path is followed. Test cases that would achieve decision coverage are typically a tiny subset of the test cases that would achieve path coverage. In practice, most non-trivial programs (and all programs with unconstrained loops, such as ‘while’ loops) have a potentially infinite number of possible paths through them and so measuring the percentage covered is practically infeasible
b) Is not correct. Coverage of business flows can be a focus of use case testing, but use cases rarely cover a single component. It may be possible to cover the decisions within business flows, but only if they were specified in enough detail, however this option only suggests coverage of “business flows” as a whole. Even if business flows would cover some decisions, the measure “Decision Coverage” don’t measure the percentage of business flows, but the percentage of decision outcomes exercised by the business flows
c) Is not correct. Achieving full decision coverage does require all ‘if’ statements to be exercised with both true and false outcomes, however, there are typically several other decision points in the code (e.g., ‘case’ statements and the code controlling loops) that also need to be taken into consideration when measuring decision coverage
d) Is correct. Decision coverage is a measure of the proportion of decision outcomes exercised (normally presented as a percentage)

Explanation / Rationale

a) Is not correct. ‘Beware of the pesticide paradox’ is concerned with re-running the same tests and their fault-finding effectiveness decreasing
b) Is not correct. This testing principle is concerned with performing testing differently based on the context (e.g., games vs safety-critical)
c) Is not correct. This testing principle is concerned with the difference between a tested and fixed system and a validated system. No ‘errors’ does not mean the system is fit for use
d) Is correct. If clusters of defects are identified (areas of the system containing more defects than average), then testing effort should be focused on these areas

Explanation / Rationale

a) Is correct. Impact analysis may be used to identify those areas of the system that will be affected by the fix, and so the extent of the impact (e.g., necessary regression testing) can be used when deciding if the change is worthwhile
b) Is not correct. Although testing migrated data is part of maintenance testing (see conversion testing), impact analysis does not identify how this is done
c) Is not correct. Impact analysis shows which parts of a system are affected by a change, so it can show the difference between different hot fixes in terms of the impact on the system, however it does not give any indication of the value of the changes to the user
d) Is not correct. Impact analysis shows which parts of a system are affected by a change; it cannot provide an indication of the effectiveness of test cases

Explanation / Rationale

a) Is correct. A crash is clearly noticeable by the user
b) Is not correct. This is a defect, not a failure since there is something wrong in the code. It may not result in a visible or noticeable failure, for example if the changes in the source code file are only in comments
c) Is not correct. The use of wrong input variables may not result in a visible or noticeable failure, for example if nobody uses this particular algorithm; or if the wrong input variable has a similar value to the correct input variable; or if the FALSE result of the algorithm is not used
d) Is not correct. This type of fault will not necessarily lead to a failure; for example, if no one uses this special algorithm

Explanation / Rationale

Given the annotated state model below:

The options achieve the following transition coverage:
a) Is not correct: OFF (2) WAIT (1) OFF (2) WAIT (3) TRICKLE (5) CHARGE (9) HIGH (10) CHARGE (7) LOW = 7 transitions (out of 10)
b) Is not correct: WAIT (3) TRICKLE (4) WAIT (1) OFF (2) WAIT (3) TRICKLE (5) CHARGE (7) LOW (8) CHARGE = 7 transitions (out of 10)
c) Is not correct: HIGH (10) CHARGE (7) LOW (8) CHARGE (6) TRICKLE (4) WAIT (3) TRICKLE (4) WAIT (3) TRICKLE = 6 transitions (out of 10)
d) Is correct: WAIT (3) TRICKLE (5) CHARGE (9) HIGH (10) CHARGE (6) TRICKLE (4) WAIT (1) OFF (2) WAIT = 8 transitions (out of 10)

Explanation / Rationale

a) Is not correct. See reason from correct answer
b) Is not correct. See reason from correct answer
c) Is correct. This is a case where the requirement gives an enumeration of discrete values. Each enumeration value is an equivalence class by itself; therefore, each will be tested when using equivalence partitioning test technique
d) Is not correct. See reason from correct answer

Explanation / Rationale

a) Is correct. Defects found early are often much cheaper to remove than defects detected later in the lifecycle
b) Is not correct. Dynamic testing still has its challenging because they find other types of defects
c) Is not correct. This is dynamic testing
d) Is not correct. Static testing is important for safety-critical computer systems

Explanation / Rationale

a) Is not correct. The root cause is the distraction that the programmer experienced while programming
b) Is not correct. The accepting of invalid inputs is the failure
c) Is not correct. The error is the mistaken thinking that resulted in putting the defect in the code
d) Is correct. The problem in the code is a defect

Explanation / Rationale

a) Is not correct. Risk is determined by considering a combination of the likelihood of problem situations and the harm that may result from them but cannot be calculated by adding these together (the probability would be in the range 0 to 1 and the harm could be in dollars)
b) Is not correct. Risk is determined by considering a combination of a likelihood and an impact. This definition only considers likelihood and chance (both forms of probability) with no consideration of the impact (or harm)
c) Is correct. See reasons from incorrect answers
d) Is not correct. Risk is determined by considering a combination of a likelihood and an impact. This definition only considers hazards and losses (a hazard is a bad event, like a risk, while loss is a form of impact) with no consideration of the likelihood (or probability)

Explanation / Rationale

a) Is correct. Exploratory testing is a form of experience-based testing, which benefits from the skills and experience of the tester
b) b) Is not correct. Exploratory testing is useful to complement formal testing techniques
c) Is not correct. In session-based test management, exploratory testing is conducted within a defined time-box, and the tester uses a test charter containing test objectives to guide the testing
d) Is not correct. Exploratory testing can incorporate the use of other black-box, white-box, and experience-based techniques referenced in this syllabus

Explanation / Rationale

a) Is not correct. This is the Glossary definition of quality assurance
b) Is correct. This is the Glossary definition of quality
c) Is not correct. This is the Glossary definition of security
d) Is not correct. This is the Glossary definition of cost of quality

Explanation / Rationale

Proposed test case covers all five possible single valid transitions in the given state diagram (S1->S2, S2->S1, S2->S3, S3->S2, and S3->S1).

Thus:
a) Is not correct. Because no invalid transitions are covered
b) Is correct. Because all valid transitions are covered
c) Is not correct. Because all valid transitions are covered
d) Is not correct. Because the test cases do not have pairs of transitions specified

Explanation / Rationale

a) Is not correct. Statement coverage measures the percentage of statements exercised by test cases
b) Is correct. Statement testing exercises the executable statements in the code. Statement coverage is measured as the number of statements executed by the tests divided by the total number of executable statements in the test object, normally expressed as a percentage
c) Is not correct. The coverage does not measure pass/fail
d) Is not correct. It is a metric and does not provide true/false statements

Explanation / Rationale

a) Is not correct. Statement coverage is a measure of the proportion of executable statements exercised by tests. The number of executable statements is often close to the number of lines of code minus the comments, but this option only talks about the number of lines of code exercised and not the proportion exercised
b) Is correct. Statement coverage is a measure of the proportion of executable statements exercised by tests (normally presented as a percentage)
c) Is not correct. Statement coverage is a measure of the percentage of executable statements exercised by tests, however many of the lines of source code are not executable (e.g., comments)
d) Is not correct. Statement coverage is a measure of the proportion of executable statements exercised by tests. This option only talks about the number of executable statements exercised and not the proportion (or percentage) exercised

Explanation / Rationale

Considering the attributes:

• There is a role of a scribe – specified for walkthroughs, technical reviews, and inspections; thus, the reviews being performed cannot be informal reviews
• The purpose is to detect potential defects – the purpose of detecting potential defects is specified for all types of review.
• The review meeting is led by the author – this is not allowed for inspections and is typically not the author for technical reviews, but is part of walkthroughs, and allowed for informal reviews
• Reviewers find potential issues by individual review - all types of reviews can include individual review (even informal reviews)
• A review report is produced - all types of reviews can produce a review report, although it would be less likely for an informal review

Thus:
a) Is not correct
b) Is correct
c) Is not correct
d) Is not correct

Explanation / Rationale

The correct pairings of examples to entry and exit criteria are:

• Entry criteria

o (3) The trading performance test environment has been designed, set-up and verified – example of the need for a test environment to be ready before testing can begin
o (5) The autopilot design specifications have been reviewed and reworked – example of the need for the test basis to be available before testing can begin
o (6) The tax rate calculation component has passed unit testing – example of the need for a test object to have met the exit criteria for a prior level of testing before testing can begin

• Exit criteria

o (1) The original testing budget of $30,000 plus contingency of $7,000 has been spent – example of spending the testing budget being a signal to stop testing
o (2) 96% of planned tests for the drawing package have been executed and the remaining tests are now out of scope – example of all the planned tests being run being a signal to stop testing (normally used alongside the exit criteria on outstanding defects remaining)
o (4) Current status is no outstanding critical defects and two high-priority ones – example of the number of outstanding defects achieving a planned limit being a signal to stop testing (normally used alongside the exit criteria on planned tests being run).

Thus:
a) Is not correct
b) Is not correct
c) Is not correct
d) Is correct

Explanation / Rationale

a) Is not correct. This is the Glossary definition of condition coverage
b) Is not correct. Decision coverage is a higher level of coverage and the two terms are not defined as synonyms in the Glossary
c) Is not correct. This is the Glossary definition of statement coverage
d) Is correct. This is the Glossary definition of coverage as applied to decisions

Explanation / Rationale

Traceability assists with:

• Selecting regression tests in terms of analyzing the impact of changes (1D)
• Evaluating completeness of test execution which makes testing auditable (2B)
• Identifying which user stories have open defect reports which improves understandability of test status reports to include status of test basis items (3A)
• Evaluating whether the number of tests for each requirement is consistent with the level of product risk which provides information to assess test process quality (i.e., alignment of test effort with risk) (4C)

Thus:
a) Is not correct
b) Is not correct
c) Is not correct
d) Is correct

Explanation / Rationale

a) Is correct. See reasons from incorrect answers
b) Is not correct. The “degree of tester’s independence” does not play a role in exit criteria
c) Is not correct. “Availability of test environment” is an entry criterion
d) Is not correct. “The Qualification of Tester” is not a typical exit criterion

Explanation / Rationale

a) Is not correct. If there was no agreement on targets, it is impossible to reach the targets. Since this situation can´t occur, this is not a scenario happening in reality
b) Is not correct. The test case is objectively wrong, since under these conditions no bonus is paid because the agreed target was not reached
c) Is not correct. There was no agreement on targets, it is impossible to reach the targets. Since this situation can´t occur, this is not a scenario happening in reality
d) Is correct. The test case describes the situation that the too short period of employment and the non-fulfilment of the agreed target leads to non-payment of the bonus. This situation can occur in practice but is missing in the decision table

Explanation / Rationale

a) Is not correct. R4 is dependent on R2, so R2 should be tested before R4
b) Is not correct. R4 is dependent on R2, R5 and R6, so R5 and R6 should be tested before R4
c) Is correct. The tests are specified in a sequence that takes the dependencies into account
d) Is not correct. R2 is dependent on R3, so R3 should be tested before R2

Explanation / Rationale

a) Is not correct. Defect clustering has to do with where defects are most likely to be found, not whether all of them can be found
b) Is correct. Testing can show the presence of defects but cannot prove their absence, which makes it impossible to know if you have caught all the bugs. Further, the impossibility of exhaustive testing makes it impossible for you to catch all the bugs
c) Is not correct. This principle says that you can find and remove many bugs but still release an unsuccessful software product, which is not what the product owner is asking you to ensure
d) Is not correct. Root cause analysis is not a testing principle

Explanation / Rationale

a) Is not correct. This is an example of a mistake made by the developer
b) Is not correct. This is an example of a defect (something wrong in the code that may cause a failure)
c) Is correct. This is a deviation from the expected functionality - a cruise control system should not be affected by the radio
d) Is not correct. This is an example of a defect (something wrong in a specification that may cause a failure if subsequently implemented)

Explanation / Rationale

a) Is not correct. Testing is context dependent, regardless of it being manual or automated (principle #6), but does not result in detecting a decreasing number of faults as described above
b) Is not correct. Exhaustive testing is impossible, regardless of the amount of effort put into testing (principle #2)
c) Is correct. Principle #5 says “If the same tests are repeated over and over again, eventually these tests no longer find any new defects. To detect new defects, existing tests and test data may need changing, and new tests may need to be written.” Automated regression testing of the same test cases will not bring new findings
d) Is not correct. ”Defect cluster together” (principle #4). A small number of modules usually contain most of the defects, but this does not mean that fewer and fewer defects will be found

Explanation / Rationale

Considering each statement:

1. Each development activity should have a corresponding testing activity. TRUE
2. Reviewing should start as soon as final versions of documents become available. FALSE – it should start as soon as drafts are available
3. The design and implementation of tests should start during the corresponding development activity. FALSE – the analysis and design of tests should start during the corresponding development activity, not the implementation
4. Testing activities should start in the early stages of the software development lifecycle. TRUE

Thus:
a) Is not correct
b) Is not correct
c) Is not correct
d) Is correct

Explanation / Rationale

a) Is not correct. System testing does not test interfaces between components and interactions between different parts of the system; this is a target of integration tests
b) Is correct. Examples of work products that can be used as a test basis for component testing include detailed design, code, data model, component specifications. Examples of work products for system testing include system and software requirement specifications (functional and non-functional) use cases
c) Is not correct. Component testing does not ONLY focus on functional characteristics
d) Is not correct. Component tests are also executed by developers, whereas system testing typically is the responsibility of (independent) testers

Explanation / Rationale

a) Is not correct. While impact analysis is useful during maintenance testing it is not necessary for confirmation testing since confirmation testing is on the intended effects of a bug fix or other change
b) Is not correct. Confirmation and regression testing are two separate activities, and confirmation testing is not part of system design
c) Is correct. Impact analysis can be used to select regression tests for maintenance testing
d) Is not correct. Confirmation testing is not part of impact analysis, though confirmation testing will typically happen during maintenance testing

Explanation / Rationale

a) Is not correct. While deviation from standards is a typical, we are not given any standard with which the user stories should comply
b) Is correct. Contradiction is a typical requirements defect. AC3 and AC5 conflict if the Rod is touched to an object that extends more than 1 meter in any direction from the point at which touched, since AC1 does not limit the size of the objects to be touched
c) Is not correct. While security vulnerabilities are typical defects, there is nothing here related to security
d) Is not correct. While test coverage gaps are typical defects, including missing tests for acceptance criteria, we are not provided with any information about which tests do and do not exist

Explanation / Rationale

a) Is not correct. Selection of tools is a test manager task
b) Is correct. See reasons from incorrect answers
c) Is not correct. The tester does not decide on the release of the test object
d) Is not correct. The tester specifies the test cases, the test manager does the prioritization

Explanation / Rationale

a) Is correct. Test management tools support the activities associated with test manager including metrics
b) Is not correct. Static code analysis metrics would have to do with the code only, not testing as a whole
c) Is not correct. These tools report on test basis coverage and code coverage only, not testing as a whole
d) Is not correct. Model-Based testing tools focus on one specific area, not testing as a whole

Explanation / Rationale

a) Is not correct. Based on definition of equivalence partition
b) Is correct. Based on definition from Glossary
c) Is not correct. Based on Glossary definition of testware
d) Is not correct. Based on definition of test oracle

Explanation / Rationale

a) Is not correct. It is described that the software architect must have completed the system specification
b) Is not correct. ‘Checklist-based’, last sentence it is documented that you should also look for defects outside the checklist
c) Is not correct. It is described: every reviewer did his review done comment
d) Is correct. It is described that a checklist is available, but who provides the checklist?

Explanation / Rationale

a) Is not correct. The test result is given in the short summary
b) Is correct. When testing different versions of software, identifying information is necessary
c) Is not correct. You are just writing the defect report; hence, the status is automatically open
d) Is not correct. This information is useful for the tester but does not need to be included in the defect report

Explanation / Rationale

a) Is not correct. Statement testing exercises the executable statements in the code, which might result in the absence of certain greetings not being tested
b) Is not correct. Unless the test charter specifically mentioned testing both the presence and the absence of each type of greeting, coverage can be difficult to assess for an exploratory test
c) Is not correct. State transition testing is useful for situations where the test object responds differently to an input depending on current conditions or previous history, but in this case the test object must decide whether the current date matches a particular milestone and thus whether to display the relevant greeting
d) Is correct. Decision testing involves test cases that follow the control flows that occur from a decision point, which in this case would be deciding whether a greeting should or should not be given

Explanation / Rationale

a) Is not correct. If inadequate developer testing were the problem, the confirmation test would not pass in step 3
b) Is not correct. The same tester who successfully performed the confirmation test in step 3 is repeating it in step 5
c) Is correct. Configuration management maintains the integrity of the software. If a test that passes in step 3 fails in step 5, then something is different between those two steps. One possible difference is the test object, the option listed here. Another possible difference is the between the development environment and the test environment, but that is not an option listed here
d) Is not correct. If the developers were not fixing the defect, the confirmation test would not pass in step 3

Explanation / Rationale

There are three equivalence partitions, with the boundaries as shown:

• Invalid too low (0.4 and below)
• Valid (0.5 to 25.0)
• Invalid too high (25.1 and above)

Thus:
a) Is not correct. Only two of the equivalence partitions are covered in this set of tests
b) Is not correct. Each of those four boundary values are included in this set of tests, but the question asked for equivalence partition coverage with minimal tests, so either 0.5 or 25.0 should be dropped
c) Is correct. Each of these three equivalence partitions are covered in this set of tests
d) Is not correct. Only one of those equivalence partitions is covered by this test