Discuss ways to ensure a survey is reliable and valid

What will be an ideal response?

Answer: Surveys can provide invaluable insights, but only if they are reliable (would produce identical results if repeated under similar conditions) and valid (measure what they're supposed to measure). To conduct a survey that generates reliable and valid results, you need to choose research participants carefully and develop an effective set of questions. To develop an effective survey questionnaire, provide clear instructions to make sure people can answer every question correctly. Don't ask for information that people can't be expected to remember, such as how many times they went grocery shopping in the past year. Keep the questionnaire short and easy to answer; don't expect people to give you more than 10 or 15 minutes of their time. Whenever possible, formulate questions to provide answers that are easy to analyze. Numbers and facts are easier to summarize than opinions, for instance. Avoid leading questions that could bias your survey. If you ask, "Do you prefer that we stay open in the evenings for customer convenience?" you'll no doubt get a "yes." Instead, ask, "What time of day do you normally do your shopping?" Avoid ambiguous descriptors such as "often" or "frequently." Such terms mean different things to different people. Avoid compound questions such as "Do you read books and magazines?" When selecting people to participate in a survey, the most critical task is getting a representative sample of the entire population in question.