Why are validity and reliability important concepts to consider when designing survey questions?

What will be an ideal response?

Survey and interview methods produce only indirect measures of attitudes and behavior, measurement problems. In particular, what is recorded on a piece of paper or an audiotape is usually not an exact, error-free measure of an object. This is particularly true when the objects are attitudes, beliefs, or self-described behavior. Random errors arise by chance or happenstance and (it is hoped) cancel one another out. A systematic error, by contrast, results when a measuring device consistently over- or underestimates a true value, as when a scale always reads two pounds less than a person’s real weight. The goal of any research design, of course, is to minimize these errors. Stated differently, our investigative procedures have to ensure validity and reliability. A valid measure produces an accurate or true picture of an object, whereas a reliable one gives consistent results (measurements) across time and users.