A survey is the means of collecting a sample. In statistics, a survey is often a question or set of questions. It is critical that the questions, like the sample itself, be free from bias. Bias in a question tends to favor one response. As an extreme example of a biased question about colors might be, "Don’t you think blue is beautiful and yellow is yucky?" An unbiased form of the question might be, "Which color do you prefer, blue or yellow?" Though careful, statisticians might be concerned about the order in which the colors are offered, so that might further simplify the question to, "What is your favorite color?" This type of question is most free from bias because it does not suggest a response and therefore does not favor a response.