Techniques for Evaluating and Iteratively Refining Prompts CT-GENAI Practice Questions (Page 5)

Questions 21–25

1. 21foundation · easy

   Which approach is an example of prompt variation for a test case generation task?

   • AIncreasing the number of test cases requested in the same prompt
   • BUsing the same prompt with a different random seed
   • CRunning the same prompt multiple times
   • DChanging the wording of the instruction while keeping the intent the same
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   What is the role of feedback loops in prompt refinement?

   • ATo automate the entire testing process
   • BTo increase the number of prompt versions
   • CTo eliminate the need for human evaluation
   • DTo use information from test results and user input to improve the prompt
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A tester is generating test cases for a login page. The prompt 'Generate test cases for the login page' produces a broad but shallow set of cases. The tester wants to increase the depth of the cases (e.g., specific input combinations) without losing the broad coverage. Which adjustment best balances specificity and generality?

   • AChange the prompt to 'Generate test cases for the login page with username 'admin' and password 'password123'.'
   • BAdd a sentence: 'Include specific input combinations, such as valid/invalid usernames and passwords, and their expected outcomes.'
   • CKeep the prompt unchanged and manually add more cases.
   • DChange the prompt to 'Generate all possible test cases for the login page.'
   Reveal Solution Discussion Select an answer first
5. 24foundation · easy

   What is the benefit of making a prompt more general in a testing context?

   • AIt makes the prompt shorter
   • BIt reduces the need for evaluation
   • CIt guarantees higher precision
   • DIt can cover a wider range of test scenarios
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A team is using an AI to generate test cases for a new feature. The first prompt produces cases that are too few and too broad. They refine the prompt to add more specific requirements. The second prompt produces more cases, but some are duplicates. The team wants to continue the iterative refinement process. What is the most appropriate next step?

   • ARun the second prompt multiple times and merge the outputs.
   • BRevert to the first prompt because it had fewer duplicates.
   • CAnalyze the duplicate cases, determine why they occur, and refine the prompt to reduce redundancy (e.g., by asking for unique cases).
   • DAccept the second prompt since it produces more cases.
   Reveal Solution Discussion Select an answer first