What does condition coverage require in software testing?

Condition coverage in software testing focuses on ensuring that every condition in a decision point within the program is evaluated to both true and false outcomes. This means that each individual condition within a boolean expression or decision must be tested to ensure that all possible outcomes of those conditions are examined, thereby identifying potential errors or edge cases.

For example, if a decision involves multiple conditions that contribute to the final outcome, condition coverage requires that tests be designed such that each condition can independently be true and false within the context of the decision. This helps to increase the chances of detecting logical errors and ensures that all branches of the code that depend on the conditions are executed during testing.

This approach is distinct from other testing methodologies that may have different focuses, like decision coverage, which looks at the overall pathways taken through decisions, or data flow testing, which is more concerned with the paths that variables take through the program. Thus, condition coverage is specifically about evaluating each individual condition, making "tests for each condition in a program decision's outcomes" the correct characterization of what condition coverage requires.