Explanation / Rationale

The ‘exhaustive testing is impossible’ principle is concerned with the fact that it is not feasible to test every possible variation of test inputs in all different circumstances, except in trivial cases. Instead, testing utilizes test techniques, test case prioritization, and risk-based testing to sample from the set of possibilities and focus testing efforts.
a) Is not correct. The principle states that it is not feasible to test everything except in trivial cases. Testing everything would require testing every possible variation of test inputs in all different circumstances, which is generally infeasible as there will be a practically infinite number. Testing every possible specified output will not address this problem as the relationship between inputs and specified outputs can be different for each test object. Sometimes there may be a practically infinite number of possible specified outputs (e.g., when there are several variables representing real numbers), whereas at other times there may be just two specified outputs, such as with a single variable that can be either true or false
b) Is not correct. The principle states that it is not feasible to test every possible variation of test inputs in all different circumstances. This is because for non-trivial systems there is a practically infinite number. Therefore, in practice, documenting all possible test input variations would be impractical as it would take an infinite length of time
c) Is not correct. Starting testing as early as possible with reviews and other static testing approaches will not address the problem of there being too many possible test cases. The ‘early testing saves time and money’ principle is concerned with fixing defects early on to prevent the occurrence of subsequent defects in derived work products, thereby reducing costs and the likelihood of failures
d) Is correct. The use of equivalence partitioning and boundary value analysis to generate test cases is one way to address the principle as these test techniques provide a systematic way to derive a finite subset of all possible test cases

Explanation / Rationale

Test TC 001 must come first, followed by TC 002, to satisfy dependencies. Afterwards, TC 003 to satisfy priority and then TC 004, followed by TC 005.
Thus:
a) Is correct
b) Is not correct
c) Is not correct
d) Is not correct

Explanation / Rationale

a) Is not correct. When a user reports a software failure, thanks to the unique identification of commits, it is possible to reassemble the files from the software version which was used by the user (as well as the corresponding versions of the test scripts) and thus reproduce the failure and locate the defect faster
b) Is not correct. If a change to the test environment causes unexpected issues during testing, configuration management allows testers to roll back to a previous version of the environment. This ensures that testing can continue without being affected by the change
c) Is not correct. Configuration management ensures that all identified documentation (e.g., requirement specifications) and software items are referenced unambiguously in test documentation (e.g., test plans)
d) Is correct. This is ensured by the defect management, not configuration management process

Explanation / Rationale

a) Is correct. The test plan may include test data requirements (as part of the test approach), but not the detailed test data for test cases. Test data is part of the test cases, not the test plan. Also, it is usually impossible to define such data when the test plan is created, because it is not exactly known what the components will look like
b) Is not correct. One of the purposes of a test plan is to help ensure that the performed test activities will meet the established criteria, by including entry criteria and exit criteria. The code coverage criteria are an example of such criteria for the component test level
c) Is not correct. Documentation templates are typical content of a test plan. This helps to facilitate communication between the stakeholders by defining a standard way of communicating or reporting
d) Is not correct. One of the purposes of a test plan is to demonstrate that testing will adhere to the existing test policy and test strategy, or to explain why the testing will deviate from them. This is an example of explaining the deviation, regarding the test levels that will be (or will not be) followed

Explanation / Rationale

There are three equivalence partitions: {…, –2, –1}, {0, 1, 2}, {3, 4, …}.
For 2-value BVA all the boundary values for all the equivalence partitions must be covered.
The boundary values are –1 (for the “temperature too low” partition), 0, 2 (for the “temperature OK” partition) and 3 (for the “temperature too high” partition).

Thus:
a) Is not correct
b) Is not correct
c) Is correct. The correct option is: –1, 0, 2, 3
d) Is not correct

Explanation / Rationale

a) Is not correct. This is a common characteristic of white-box test techniques. Test conditions, test cases, and test data are derived from a test basis that may include code, software architecture, detailed design, or any other source of information regarding the structure of the software.
b) Is not correct. This is a common characteristic of white-box test techniques. Coverage is measured based on the items tested within a selected structure and the technique applied to the test basis
c) Is correct. This is a common characteristic of experience-based test techniques. This knowledge and experience include expected use of the software, its environment, likely defects, and the distribution of those defects is used to define tests
d) Is not correct. This is a common characteristic of black-box test techniques. Test cases may be used to detect gaps within requirements and the implementation of the requirements, as well as deviations from the requirements

Explanation / Rationale

Considering the attributes:

• Specified for walkthroughs, technical reviews, and inspections; thus, the reviews being performed cannot be informal reviews
• The purpose of evaluating quality is one of the most important objectives of a walkthrough
• This is not allowed for inspections and is typically not done in technical reviews. A moderator is needed in walkthroughs and is allowed for informal reviews
• All types of reviews can include individual preparation (even informal reviews)
• All types of reviews can produce a review report, although informal reviews do not require documentation

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

Explanation / Rationale

a) Is correct. Test-Driven Development (TDD) is a well-known example of a test-first approach to development
b) Is not correct. Coverage-Driven Development is not a correct example of a test-first approach to development
c) Is not correct. Quality-Driven Development is not a correct example of a test-first approach to development
d) Is not correct. Feature-Driven Development is not an example of a test-first approach to development, but is, instead, an agile software development methodology based around delivering features (as opposed to user stories in Scrum)

Explanation / Rationale

a) Is not correct. Quality should be the responsibility of everyone working on the project and not the sole responsibility of the test team
b) Is not correct. First, it is not a benefit if an external test team does not meet delivery deadlines, and second, there is no reason to believe that external test teams will feel they do not have to meet strict delivery deadlines
c) Is not correct. It is bad practice for the test team to work in complete isolation, and we would expect an external test team to be concerned with changing project requirements and communicating well with developers
d) Is correct. Specifications are never perfect, meaning that assumptions will have to be made by the developer. An independent tester is useful in that they can challenge and verify the assumptions and subsequent interpretation made by the developer

Explanation / Rationale

Considering:

• The test basis for acceptance testing is the user’s business needs (1D)
• Communication between components is tested during component integration testing (2B)
• Failures in logic can be found during component testing (3A)
• Business rules are the test basis for system testing (4C)

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

Explanation / Rationale

In branch testing the coverage items are branches, which are represented by the edges of a control flow graph. There are 8 edges in the control flow graph.
Thus:
a) Is not correct
b) Is not correct
c) Is correct
d) Is not correct

Explanation / Rationale

a) Is not correct. In the whole-team approach, testers play a vital role by sharing their testing expertise with the team and guiding product development. They collaborate with other team members to achieve the desired quality levels and work with business representatives to create acceptance tests. Testers also partner with developers to determine the optimal test strategy and automation approaches
b) Is correct. By leveraging the diverse skill sets of each team member most effectively, the whole-team approach fosters superior team dynamics, promotes robust communication and collaboration, and generates a synergistic effect that benefits the entire project
c) Is not correct. The whole-team approach allows any team member with the requisite skills and knowledge to undertake any task, thus specialist team members are not an advantage of this approach
d) Is not correct. There is no specific guidance on the optimum size of teams using the whole-team approach, and there is no suggestion that larger teams are better

Explanation / Rationale

a) Is correct. Obtaining feedback from stakeholders early and often in the software development process can be highly beneficial. It facilitates early communication of potential quality issues, can prevent misunderstandings about requirements, and ensures that any changes in stakeholder requirements are understood and implemented sooner
b) Is not correct. The feedback is from stakeholders, and them providing feedback is unlikely to improve their understanding of their own user requirements
c) Is not correct. Obtaining feedback from stakeholders early and often in the software development process can be highly beneficial. It facilitates early communication of potential quality issues, can prevent misunderstandings about requirements, and ensures that any changes in stakeholder requirements are understood and implemented sooner. However, because changes in requirements can be understood and implemented sooner, it does not mean that unlimited changes to requirements are encouraged
d) Is not correct. The feedback is from stakeholders and does not cover communication to them. Communications to end users could include them being told about which requirements will not be implemented prior to release, but ideally this should not happen at all

Explanation / Rationale

a) Is not correct. Testers should review work products as soon as drafts are available to enable early testing as part of a shift-left approach. If they waited until the next development phase, then unnecessary development (and test) work could be started on unreviewed, flawed work products
b) Is correct. Testers should review work products as soon as drafts are available to enable early testing as part of a shift-left approach
c) Is not correct. Testers typically review work products that form the test basis as part of test analysis, not before test analysis and design
d) Is not correct. Testers should review work products as soon as drafts are available to enable early testing as part of a shift-left approach. Waiting until they are published means that any defects that could be found by tester review will be in the published document

Explanation / Rationale

a) Is correct. Test automation can provide measures that are too complicated for humans to derive, such as white-box test coverage measures for all but the most trivial code
b) Is not correct. By using test tools the responsibility for the testing is NOT shared with the tool vendor as the vendor is not involved in the testing, and it is the tester’s responsibility. The only possible responsibility that could be assigned to the tool vendor is if the tool fails to work as expected and provides incorrect test results
c) Is not correct. Testers still need to perform critical thinking when analyzing anomalies in the test results to determine their likely cause
d) Is not correct. Neither testers nor tools can generate test cases simply from an analysis of the program code as the code is the implementation and provides no information on the expected results, which will need to come from another part of the test basis, such as the design specification

Explanation / Rationale

a) Is not correct. In agile software development, deliverables are produced in each iteration, and the frequent delivery of increments necessitates extensive regression testing. Although some (or all) of this regression testing may be automated, the regression testing (automated or not) cannot be replaced by system test automation
b) Is correct. If a sequential development model is used, then early in the lifecycle no code is available for execution, and so during this time static testing (e.g., reviews) is performed. Later in the lifecycle, when code is available for execution, dynamic testing is possible. Note, however, that preparation for dynamic testing will often occur early in any software development lifecycle
c) Is not correct. If an iterative development model, like agile software development, is used, then component tests may well be used for regression testing for each iteration. In which case, there is a strong argument for automating these component tests, which will have to be run frequently, and there is unlikely to be a strong argument for developers performing these component tests manually
d) Is not correct. In most incremental development models, deliverables are produced in each increment, requiring both static and dynamic testing at all test levels for each increment delivered

Explanation / Rationale

a) Is not correct. Component Test-Driven Development is not a correct example of a test-first approach to development
b) Is not correct. Integration Test-Driven Development is not a correct example of a test-first approach to development
c) Is not correct. System Test-Driven Development is not a correct example of a test-first approach to development
d) Is correct. Acceptance Test-Driven Development (ATDD) is a well-known example of a test-first approach to development

Explanation / Rationale

a) Is correct. A fundamental strength that all white-box test techniques share is that the entire software implementation is taken into account during testing, which facilitates defect detection even when the software specification is vague, outdated or incomplete. This means white-box testing can find defects such as an extra feature added to the code (either accidentally or deliberately) that is not supposed to be there, which black-box testing cannot detect
b) Is not correct. The fact that the coverage can be precisely defined is not the right reason. The achieved level of coverage would have much more impact than the possibility to measure the coverage
c) Is not correct. If the software does not implement one or more requirements, white-box testing is unlikely to detect the resulting defects of omission
d) Is not correct. While this is true, this is not the right answer, because there is no connection between the capability to be used in both static testing and dynamic testing and the claim that white-box testing facilitates defect detection with poor specifications

Explanation / Rationale

a) Is not correct. Collaborative user story writing means that all stakeholders create the user stories collaboratively, to obtain the shared vision
b) Is correct. Collaborative user story writing means that all stakeholders create the user stories collaboratively, to obtain the shared vision
c) Is not correct. Collaborative user story writing means that all stakeholders create the user stories collaboratively, to obtain the shared vision
d) Is not correct. This is the list of properties that each user story should have, not the description of the collaboration-based approach

Explanation / Rationale

a) Is not correct. Exploratory testing uses test charters, not a list of possible defects/failures. Although exploratory testing can incorporate the use of other test techniques, in this case fault attack is the most likely option
b) Is correct. This is a list of possible failures. Fault attacks are a methodical approach to the implementation of error guessing and require the tester to create or acquire a list of possible errors, defects and failures, and to design tests that will identify defects associated with the errors, expose the defects, or cause the failures
c) Is not correct. The tester is using a checklist of items to support their testing. Both error guessing and checklist-based testing use such lists, however, the list here is of possible failures, not test conditions, and so the MOST PROBABLE test technique is fault attack, which focuses on errors, defects and failures
d) Is not correct. BVA is based on an analysis of boundary values of equivalence partitions. The above list does not mention equivalence partitions or their boundaries

Explanation / Rationale

a) Is not correct. This sequence of five events covers 4 different valid transitions (both “NotAvailable” events correspond to the same transition between S1 and S3). This test case covers 4 out of 7 valid transitions
b) Is correct. This sequence of five events covers 5 different transitions (the first “Available” event corresponds to a transition between S1 and S2, and the second “Available” event corresponds to a transition between S3 and S2, so two different transitions are covered). This test case covers 5 out of 7 valid transitions and achieves the highest valid transitions coverage
c) Is not correct. This sequence of five events covers 3 different transitions (both “Available” events correspond to the same transition from S1 to S2; both “ChangeRoom” events correspond to the same transition from S2 to S1). This test case covers 3 out of 7 valid transitions
d) Is not correct. This sequence of five events does not represent a feasible test case, because after “Cancel” the system ends up in the End state and no further valid transitions can be executed

Explanation / Rationale

a) Is not correct. The miscalculation of bonuses is a failure by the system, not a defect
b) Is not correct. The system not suitably supporting disabled users is a failure which eventually results in a fine, but the fine itself is not a failure (it appears to be the correct functioning of the regulatory system)
c) Is correct. The error is made by the programmer and this mistake is caused by them working under severe time pressure, which is the root cause of the subsequent defect
d) Is not correct. The poor design of the user interface, which does not suitably address disabled users, is a design defect caused by the designer error. Thus the design of the user interface includes a design defect not a designer error

Explanation / Rationale

a) Is not correct. Dynamic testing does cause failures (from which defects can then be located and fixed). However, debugging is concerned with locating defects and fixing these defects. Therefore, debugging does not fix failures
b) Is not correct. Both testing and debugging contribute to improving the quality of the test object, so should really both be considered positively. Debugging is generally considered to be a positive activity as it is fixing something. Dynamic testing does involve intentionally causing the test object to fail, which is why some people consider it a negative activity, but that is a very narrow view (and not one typically held by testers). Both positive and negative test cases are possible. Positive test cases check that the test object correctly performs what it is supposed to do, while negative testing checks that the test object does not do what it is not supposed to do
c) Is correct. Testing determines that defects exist either directly through observation of the defect in reviews (or by a tool in static analysis), or indirectly by causing a failure in dynamic testing. Debugging is a separate activity from testing (normally performed by developers) and is concerned with locating defects (only for dynamic testing) and fixing the defects
d) Is not correct. The causes of defects are typically human errors. Testing finds defects either directly through static testing, or indirectly by causing failures in dynamic testing, and debugging fixes defects. So, testing does not find the cause of defects and debugging does not fix the causes of defects

Explanation / Rationale

a) Is not correct. It is impossible to prove that there are no defects anymore in the system under test. See testing principle 1
b) Is not correct. See testing principle 7
c) Is correct. Testing finds defects and failures which reduces the level of risk and at the same time gives more confidence in the quality level of the test object
d) Is not correct. It is impossible to test all combinations of inputs (see testing principle 2)

Explanation / Rationale

The following three transitions:

“REQUESTING -> CONFIRMED”
“WAITING LIST -> CONFIRMED”
“WAITING LIST -> END”

cannot appear in the same test case, which suggests that at least three test cases are required. All the other transitions can appear in combination with one or more of these three transitions, so we need a minimum of three test cases. In fact, only three sequences are possible:

TC1: START (Room request) → REQUESTING (Available) → CONFIRMED (Pay) → END
TC2: START (Room request) → REQUESTING (Not available) → WAITING LIST (Available) → CONFIRMED (Pay) → END
TC3: START (Room request) → REQUESTING (Not available) → WAITING LIST (Cancel) → END

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

Explanation / Rationale

Consider:

i. It is a Project risk
ii. It is a Product risk
iii. It is a Product risk
iv. It is a Project risk
v. It is a Product risk

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

Explanation / Rationale

a) Is correct. It is important that testers are involved from the beginning of the software development lifecycle (SDLC). It will increase understanding of design decisions and will detect defects early.
b) Is not correct. Both developers and testers will have more understanding of each other's work products and how to test the code
c) Is not correct. End users will not help the testers in increasing the quality of defect reports; also, users usually do not participate in low-level testing levels like integration testing
d) Is not correct. Being certified does not automatically mean that the tester will be better in test design

Explanation / Rationale

Given the following description of test analysis:
To identify the features that require testing, the test basis is analyzed and defined as test conditions, which are then prioritized along with related risks. The systematic identification of test conditions as coverage items often involves using test techniques both during test analysis and as part of the test design activity.
From the above description, it can be seen that test techniques are often used in the test analysis and test design activities. Boundary value analysis and equivalence partitioning are test techniques.
a) Is not correct. Test implementation is not likely to involve the use of test techniques as it is mostly concerned with assembling test cases into test procedures, while test techniques create test cases
b) Is correct. Test design is likely to involve the use of test techniques to create test cases from test conditions and coverage items
c) Is not correct. Test execution is not likely to involve the use of test techniques as it is mostly concerned with executing test procedures (and so test cases), while test techniques create test cases
d) Is not correct. Test monitoring is not likely to involve the use of test techniques. Test monitoring is mostly concerned with ongoing checks to ensure the plan is being followed, while test techniques create test cases
e) Is correct. Test analysis is likely to involve the use of test techniques to identify test conditions

Explanation / Rationale

“test” and “error” transitions cannot occur in one test case.
Neither can both “done” transitions.
This means we need at least three test cases to achieve transition coverage. For example:

TC1: test, done
TC2: run, error, done
TC3: run, pause, resume, pause, done

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

Explanation / Rationale

Risk assessment can use a quantitative or qualitative approach, or a mix of them. In the quantitative approach the risk level is calculated as the multiplication of risk likelihood and risk impact. So, Risk level = Risk likelihood * Risk impact
Then, Risk impact = Risk level / Risk likelihood.
In our case, Risk impact = $1,000 / 50% = $1,000 / 0.5 = $2,000.
Thus:
a) Is not correct
b) Is correct
c) Is not correct
d) Is not correct

Explanation / Rationale

a) Is not correct. Test environment readiness is a resource availability criterion; hence it belongs to the entry criteria
b) Is not correct. This is a resource availability criterion; hence it belongs to the entry criteria
c) Is correct. Estimated defect density is a measure of diligence; hence it belongs to the exit criteria.
d) Is not correct. Requirements translated into a given format result in testable requirements; hence it belongs to the entry criteria
e) Is correct. Automation of regression tests is a completion criterion; hence it belongs to the exit criteria

Explanation / Rationale

a) Is not correct. A line with a defect, when executed, does not have to cause a failure. For example, a line x := y / z will cause a failure only when z equals 0
b) Is not correct. 100% statement coverage does not guarantee 100% branch coverage. For example, a test case with x=0 for the code

1. IF (x=0) THEN
2. A;
3. ENDIF

achieves 100% statement coverage but does not cover the branch from 1 to 3
c) Is correct. 100% statement coverage means that each executable statement was executed at least once
d) Is not correct. The removed test case may provide coverage of some statements that are not covered by either of the other two test cases, in which case the remaining two test cases together will not achieve 100% statement coverage

Explanation / Rationale

a) Is not correct. It is often possible to use dynamic testing to cause a test object to fail in ways that could never be achieved by the users, such as by using fault injection. However, if the failure can never occur with real end users, then identifying it is not especially valuable as testing is ultimately aimed at improving the work product for the end users. Spending time testing for failures that cannot occur with real users is not an efficient use of a tester’s time
b) Is not correct. Static testing in the form of static analysis is used by developers to identify defects in their program code earlier than can be achieved through dynamic testing. Note, however, that static testing (and static analysis) is used to detect defects, not failures, which are found by dynamic testing. Thus it is the use of the term ‘failures’ that makes this an incorrect option
c) Is not correct. Static analysis directly detects anomalies in code, which may be defects, and this is normally for the developer, not the customer. The provision of evidence for release by the use of static analysis of elements that provide no output is nonsense
d) Is correct. Reviews are a form of static testing that can be applied from the very start of the software development lifecycle and are used to find defects that can be removed before subsequent development activities waste effort on faulty requirements. If the defects are not detected and removed early on, then when the defect is found derived work products, such as the design and code, will need to be changed as they were based on faulty requirements

Explanation / Rationale

a) Is not correct. Validating that documented requirements are met is incorrect as validation is concerned with meeting user requirements and expectations, while verification is concerned with meeting specified requirements, so this would be correct if we replaced ‘validating’ with ‘verifying’
b) Is correct. Causing failures and identifying defects is probably the most common objective of dynamic testing
c) Is not correct. Initiating errors and identifying root causes is incorrect because testers do not initiate errors, they try to cause failures. Errors are typically made by developers (and cannot really be initiated) and result in defects, which testers attempt to identify either directly through static testing or indirectly through failures with dynamic testing. Identifying root causes is useful but is part of debugging, which is a separate activity to testing
d) Is not correct. Verifying the test object meets user expectations is incorrect as verification is concerned with checking specified (documented) requirements are met, while validation is concerned with meeting user requirements and expectations, so this would be correct if we replaced ‘verifying’ with ‘validating’

Explanation / Rationale

a) Is correct. The number of defects found is related to the test object quality
b) Is not correct. This is the measure of the test efficiency not the test object quality
c) Is not correct. The number of test cases executed does not tell us anything about the quality; test results might do
d) Is correct. defect density is related to the test object quality
e) Is not correct. Time to repair is a process metric. It does not tell us anything about the product quality

Explanation / Rationale

From the graph we have:

A(4)=6 and A(3)=8 (the last two gray boxes).

From the formula we obtain:

E(5) = (3*A(4) + A(3)) / 4 = (3*6+8) / 4 = 26 / 4 = 6.5 person-days.

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

Explanation / Rationale

Considering each of the listed review types:

1. Technical review - This type of review is performed by technically qualified reviewers and led by a moderator. The objectives are to gain consensus and make decisions on technical problems while also evaluating quality and building confidence in the work product, generating new ideas, motivating and enabling authors to improve, and detecting anomalies
2. Informal review - The main objective is to detect anomalies. The process is not defined and does not require formal documented output
3. Inspection - This is the most formal review type, and it follows the complete generic review process. The primary objective is to find the most anomalies, and other objectives include evaluating quality and building confidence in the work product, motivating and enabling authors to improve, and collecting metrics that can be used to enhance the software development life cycle (SDLC), including the inspection process. The author cannot act as the review leader or scribe
4. Walkthrough - Led by the author, this type of review serves various objectives such as evaluating quality and building confidence in the work product, educating reviewers, gaining consensus, generating new ideas, motivating and enabling authors to improve, and detecting anomalies. Reviewers might perform an individual review before the walkthrough, but this is not mandatory

A. Includes objectives such as gaining consensus, generating new ideas, and motivating authors to improve (A)
B. The main objective is detecting potential defects and it generates no formal documented output (D)
C. The main objective is detecting potential defects and it requires metrics collection to support process improvement (C)
D. Includes objectives such as educating reviewers, gaining consensus, generating new ideas and detecting potential defects (B)

Thus:
a) Is not correct
b) Is correct. The correct match is: 1A, 2D, 3C, 4B
c) Is not correct
d) Is not correct

Explanation / Rationale

a) Is not correct. The detection of additional high-severity defects would be a benefit of test automation, rather than a risk
b) Is not correct. The provision of measures that are too complicated for humans to derive themselves is normally considered to be a benefit of test automation
c) Is correct. If the test automation is incompatible with the development platform, then it will not be able to integrate them, and, for instance, pass test inputs to the test object and receive test results from the test object
d) Is not correct. Substantially reduced test execution times would normally be considered a benefit that is provided by test automation

Explanation / Rationale

a) Is not correct. The book provides general guidance, and is not a formal requirements document, a specification, or a set of use cases, user stories, or business processes
b) Is not correct. While you could consider the list as a set of test charters, it more closely resembles the list of test conditions to be checked
c) Is correct. The list of user interface best practices is the list of test conditions to be systematically checked
d) Is not correct. The tests are not focused on failures that could occur, but rather on knowledge about what is important for the user, in terms of usability

Explanation / Rationale

a) Is not correct. It is more often used in behavior-driven development (BDD)
b) Is not correct. It is the description of test-driven development (TDD)
c) Is correct. In acceptance test-driven development (ATDD) tests are written from acceptance criteria as part of the design process
d) Is not correct. It is used in BDD