Which of the following best describes 'validity' in measurement?

Validity in measurement fundamentally refers to the degree to which a test accurately assesses what it is intended to measure. This means that if a test claims to measure a specific psychological trait, such as anxiety, its validity would be determined by how well it truly reflects that trait rather than some other unrelated concept.

When considering other options, consistency over time relates more to the concept of reliability, which is about obtaining stable and consistent results across various administrations of the test. The ease of replicating results pertains to research replicability, which is also tied to reliability rather than validity. Randomness in sampling refers to selecting participants randomly to ensure the sample represents the larger population, and while important for the overall study design, it does not directly address whether a measurement tool effectively measures its intended content.

Thus, option B clearly encapsulates the essence of validity—measuring what the test claims to measure.