A poorly written and worded question that has different meanings to different respondents, when used in a marketing research setting, is said to have poor
A. validity
B. accuracy
C. reliability
D. confidence interval
E. sample
Answer: (A) Validity is the extent that a research instrument (such as a questionnaire) measures what it purports, or is intended, to measure. If the question has a different meaning to different people, it is invalid. (B) is incorrect. Accuracy refers to the extent that collected data are error free. A poorly worded question may not have an impact on accuracy. Reliability (C) is incorrect. Reliability refers to the fact that the questionnaire, or research instrument, measures the same thing over and over. Thus, every time the instrument is used it measures the same thing. The research can be reliable; however, it still may not be valid. In other words it is measuring the incorrect thing over and over. (D) is incorrect. Confidence intervals have to do with sampling accuracy. A confidence interval is the range on either side of an estimate that would likely contain the true population values. Samples (E) have nothing to do with question wording, and is thus incorrect. Samples are small portions of a population under study that should reflect the overall population.