Black-box testing focuses on

Software testing techniques are classified by how much of a program's internal implementation the tester can see. Black-box testing (also called functional or behavioral testing) treats the program as a closed box: test cases are derived purely from its specification, and the tester has no visibility into the source code, algorithms, or internal data structures. Testing that instead examines internal logic, statements, and control-flow paths falls under white-box (structural) testing.

Applying this to the question: a black-box tester supplies inputs to the program per its requirements and observes only the outputs it returns, comparing them against the expected results. Pass or fail is decided entirely from this input-output comparison — how the program internally computes the result is irrelevant to the test.

• Internal code structure — the arrangement of statements, modules, and control-flow paths — is inspected by white-box (structural) testing, not black-box testing.

• Algorithms — the step-by-step logic a program implements — are likewise analyzed under white-box techniques such as basis-path testing, which require visibility into the code.

• Memory management — how a program allocates and releases memory at runtime — is a resource-level concern addressed by specialized techniques such as performance or stress testing, or by code-level profiling, not by black-box functional testing.

Since black-box testing verifies a program purely by comparing supplied inputs against produced outputs, without any reference to the internal implementation, the correct answer is Input-output behavior.