ABSTRACT
Much knowledge of human behavior is based on the reports that research participants provide in response to a researcher's questions. From testing social science theories to determining the nation's unemployment rate, and from assessing public opinion to predicting future health care needs, researchers rely on the verbal reports that respondents provide in surveys. To gain an understanding of human cognition and judgment, we expose participants in psychological experiments to carefully constructed reasoning tasks. To reduce unwanted variation and to render research participants' answers comparable, the communication in these settings is highly standardized. Interviewers are trained to read the questions verbatim and to provide standardized explanations, if any, should a respondent ask for clarification. Similarly, experimenters are trained to follow a tightly designed script, even under conditions where their utterances are to appear as spontaneous. Moreover, research participants are expected to provide their answers in a predefined format, by checking a number on a rating scale or by endorsing one of several response alternatives presented to them.
