CTAL-TTA

Free CTAL-TTA practice questions

10 questions from our CTAL-TTA practice tests, each with the correct answer and a written explanation. Read them with the answers shown, or switch to quiz mode and answer first.

Back to the CTAL-TTA practice exams

Question 1 of 10, White-Box Test Techniques

Which statement correctly describes the relationship between statement coverage and decision coverage?

  1. AThey are equivalent for structured code, because every statement lies on some decision outcome
  2. BThe two measures are independent, so neither level of coverage implies anything about the other
  3. CAchieving 100% decision coverage guarantees 100% statement coverage, but not the reverse Correct answer
  4. DAchieving 100% statement coverage guarantees 100% decision coverage, but not the reverse

Answer and explanation: Decision testing subsumes statement testing: exercising every decision outcome executes every statement, while full statement coverage can leave a false outcome with no statements (such as an IF without ELSE) untested. The reverse claim is the classic misconception, the measures are not independent because of this subsumption, and they are not equivalent because an empty decision outcome contains no statements.

Question 2 of 10, White-Box Test Techniques

A safety-critical module contains the decision: IF (A AND B AND C) THEN ... The conditions are independent and the compiler does not short-circuit. What is the minimum number of test cases needed to achieve 100% modified condition/decision coverage (MC/DC) for this decision?

  1. A8 test cases
  2. B6 test cases
  3. C4 test cases Correct answer
  4. D3 test cases

Answer and explanation: For a decision with N independent conditions, MC/DC can normally be achieved with N+1 test cases; for A AND B AND C this is one all-true case plus three cases where exactly one condition is false, so 4. Three test cases cannot show the independent effect of every condition plus the true outcome, six is not a minimum, and eight is the number of combinations required for multiple condition coverage.

Question 3 of 10, Static and Dynamic Analysis

A team runs a static analysis tool on a 15-year-old code base for the first time and receives 2,000 warnings, many of them low priority. Developers threaten to ignore the tool. What is the BEST approach?

  1. ARequire all 2,000 warnings to be fixed before the next release, so that the code base starts from a clean baseline
  2. BTune the rule set, prioritise the critical warnings, and require that new or changed code introduces no new warnings Correct answer
  3. CReplace static analysis with additional system tests, because warnings that do not cause failures have no value
  4. DStop using the tool on legacy code, because static analysis is only effective for code written to modern standards

Answer and explanation: Introducing static analysis on legacy code works best when rules are tuned to reduce noise, critical issues are addressed first and a baseline prevents new violations in new or changed code. Fixing everything at once is usually unrealistic and risky, abandoning the tool loses its maintainability benefits, and system tests cannot reveal issues such as standard violations or latent defects that static analysis finds.

Question 4 of 10, Static and Dynamic Analysis

A long-running Java service slowly consumes more heap memory over several days and is eventually restarted by the operations team. Garbage collection is enabled. Which explanation and detection approach are MOST appropriate?

  1. AThe database is undersized, so a capacity test that increases the number of concurrent users should be used
  2. BObjects remain referenced, for example in a growing cache, so dynamic analysis of heap usage over time should be used Correct answer
  3. CGarbage collection prevents memory leaks entirely, so static analysis of the code for unreachable statements should be used
  4. DThe service contains a wild pointer, so a code review focused on pointer arithmetic in Java classes should be used

Answer and explanation: Garbage collection only frees unreferenced objects, so a leak still occurs when references are unintentionally retained, and dynamic analysis tools that monitor heap usage over time reveal it. Garbage collection does not eliminate this kind of leak, Java has no pointer arithmetic for wild pointers, and a capacity test of the database does not explain gradual heap growth in the service.

Question 5 of 10, Quality Characteristics for Technical Testing

In a retail system, a change to the tax calculation rules required modifications in 12 other modules because tax logic was spread across the code base. Which maintainability sub-characteristic does this MOST clearly reveal as weak?

  1. AInstallability of the new release
  2. BAnalysability of the log output
  3. CModularity Correct answer
  4. DInteroperability with tax services

Answer and explanation: Modularity is the degree to which a system is composed of discrete components so that a change to one has minimal impact on others; tax logic scattered across 12 modules shows poor modularity. Analysability concerns diagnosing failures, installability is a portability sub-characteristic, and interoperability is a compatibility sub-characteristic, none of which explains the wide impact of the change.

Question 6 of 10, Quality Characteristics for Technical Testing. Choose 2.

A Technical Test Analyst is planning tests of non-functional quality characteristics for a new trading system. Which TWO items are general planning considerations for technical testing? (Choose two.)

  1. AEnsuring the test environment is representative enough of production for valid results Correct answer
  2. BScheduling usability sessions with representative end users of the trading screens
  3. CDeriving test cases from equivalence partitions of the business input fields
  4. DWriting the user acceptance criteria together with the business stakeholders
  5. EIdentifying the tools needed and the time required to acquire them and train testers Correct answer

Answer and explanation: Planning technical tests requires attention to required tool acquisition and training and to a production-representative test environment, since non-functional results are often invalid otherwise. Equivalence partitioning of business inputs, usability sessions and user acceptance criteria are activities of the Test Analyst or business stakeholders rather than general planning considerations for technical quality characteristics.

Question 7 of 10, Quality Characteristics for Technical Testing

An embedded controller must run its control loop using no more than 60% of its CPU and 48 MB of RAM so that diagnostics can run in parallel. Which performance efficiency sub-characteristic do the tests address?

  1. ATime behaviour of the individual control loop steps
  2. BCo-existence with the diagnostic software on the controller
  3. CCapacity in terms of the maximum number of connected sensors
  4. DResource utilization Correct answer

Answer and explanation: Resource utilization is about the amounts and types of resources, such as CPU and memory, used by the product when performing its functions. Time behaviour concerns response and processing times, capacity concerns maximum limits such as connected devices, and co-existence is a compatibility sub-characteristic about sharing an environment rather than the performance efficiency measure stated in the requirement.

Question 8 of 10, Reviews

A payroll function sums monthly salaries into a 16-bit signed integer and then divides the total by the number of employees using integer division. Which code review checklist category would MOST directly lead a reviewer to find these problems?

  1. ADocumentation and comment accuracy
  2. BDefensive programming against invalid input
  3. CArithmetic operations Correct answer
  4. DCode structure and module organisation

Answer and explanation: Arithmetic operations checks cover overflow, truncation in integer division, loss of precision and rounding, which are exactly the risks in summing salaries into a small integer and dividing with integer arithmetic. Structure is about organisation and duplication, documentation concerns comments, and defensive programming concerns validating inputs and handling errors rather than the correctness of the calculation itself.

Question 9 of 10, Test Tools and Automation

A performance test script recorded with a load tool works when replayed by one virtual user, but fails for most users at 200 virtual users with 'invalid session' errors. Each login returns a unique session token. What is the MOST likely cause and fix?

  1. AThe token was recorded as a fixed value; it must be extracted from each response and reused dynamically Correct answer
  2. BThe server is at maximum capacity; the test proves that 200 users are beyond the supported concurrency limit
  3. CThe think time is too long; it should be removed so that sessions do not expire during the scripted flow
  4. DThe tool cannot generate more than one user at a time; real users must be used for the 200-user test

Answer and explanation: Recorded scripts often contain dynamic values such as session tokens, which must be correlated by extracting them from server responses and reusing them, or the server will reject replayed requests. Load tools can simulate many users, removing think time makes the load unrealistic and does not fix invalid tokens, and concluding a capacity limit is premature when the errors come from the script.

Question 10 of 10, Test Tools and Automation

A team must simulate 20,000 concurrent users of a web application but has limited load generator hardware. It is considering protocol-level virtual users instead of virtual users that drive real browsers. What is the main trade-off?

  1. AProtocol-level users measure the full browser experience but require much more hardware per virtual user
  2. BProtocol-level users need no scripting at all, but they can only simulate a maximum of 100 users in total
  3. CProtocol-level users need far fewer resources but do not measure client-side rendering time Correct answer
  4. DProtocol-level users can only be used for functional testing, so they cannot generate load on the server

Answer and explanation: Protocol-level virtual users send requests directly and are lightweight, so many can run on limited hardware, but they do not capture browser rendering and client-side script time. Browser-based users give the full experience at a higher resource cost, so the claim is reversed, protocol-level users are designed for load generation, and they require scripting and scale far beyond 100 users.