Читать книгу The Research Experience - Ann Sloan Devlin - Страница 154

Like a field study, a field experiment takes place in natural settings. Unlike the field study, a field experiment manipulates variables. Ultimately, the field experiment lacks the extent of control available in the lab, but every effort is taken to approximate that control. An example of a field experiment where more control might be possible is research in education settings, where you might have students receive different versions of instructions (by random assignment) to the same exam. You have manipulated a variable (the instructions), but the students in the classroom otherwise are in a setting that is natural and where most aspects of the setting (the instructor, the room, the exam content itself) remain controlled.

Another example of a field experiment in the world beyond the classroom is a study by Wesley Schultz et al. (2016). In this study, the focus was water conservation (in California) and the extent to which that could be influenced (measured by water usage as the dependent variable) by receiving environmental messages (Web based or postal) in people with different levels of personal norms about conserving water. There were seven conditions in the study, with three different kinds of messages that were delivered to residents: (1) descriptive norm—your water use compared with that of similar others in the neighborhood; (2) descriptive norm + injunctive—your water use and neighborhood norms again plus social approval (happy face; below neighborhood norm use of water) or disapproval (sad face; above neighborhood norm use of water) on the feedback; and (3) only information—water conservation tips. The message types were delivered in either of two forms (postal or Web based); that gives us six conditions so far. One randomized control received no information about the study (the seventh condition). In this study participants were clearly in their natural environments, their homes (in San Diego), but there was a fairly elaborate experimental manipulation to investigate how water conservation might be impacted by different kinds of feedback.