Field Experiment in Consumer Behavior Research: Definitions, Manipulation Methods, and Suggestions

Field experiment is an experimental research method corresponding to lab experiment. It has received continuous attention in recent years because of its high realism and external validity. However, past research did not reach consensus on the definition of field experiment, which puzzled the readers to understand the meaning of field experiment and to carry out a real field experiment. This research gap motivated the current paper.First, we collect previous scholars’ definitions of field experiment. Based on these past definitions, we put forward a relatively clear definition of and a criterion for field experiment. In the same part, we further distinguish the definitions of field experiment between those of some other related terms such as lab experiment, quasi experiment, field study and realistic experiment in-the-field. This part is followed by introducing two subcategories of field experiment, namely, independent field experiment and collaborative field experiment, and discussing the advantages and disadvantages of field experiment. We argue that although field experiment has disadvantages in uncovering the underlying mechanism, controlling noise, analyzing data, etc, it can improve the realism and external validity of a scholarly work. Importantly, it can increase the publication likelihood of the scholarly work to some extent.Next, we analyze a total of 108 articles which include at least one field experiment in four leading marketing journals, i.e. JCR, JMR, JM and JCP. The time span of these articles ranges from August 2014 to November 2019. Based on the summary of these 108 field experiment articles, we elaborate on the general methods of IV manipulation and DV observation. Specifically, we divide IVs of these 108 articles into several categories: individual characteristics, product stimuli, environmental situations, and other categories, etc. These IVs are mainly manipulated through natural display, actual recording and language highlighting. Regarding DVs, we find that the DVs of these 108 articles are generally observed by adopting the “pay-what-you-want” paradigm, recording participants’ actual choices, recording participants’ food intake quantity and product purchase quantity, counting website visits and clicks, and recording relevant behaviors such as information sharing and information collection, etc.In the following part, we provide a few suggestions on how to conduct two different types of field experiment (independent field experiment and collaborative field experiment). For independent field experiment, researchers can consider designing experiment flexibly, placing extra emphasis on increasing the authenticity of experiment, and avoiding participants’ awareness. While for collaborative field experiment, researchers can consider using a large sample size, understanding the needs of non-academic partners, and bringing practical benefits to non-academic partners, etc. Sometimes researchers also need to meet the anonymity requirements of non-academic partners.In the final part, we discuss when researchers should carry out field experiment and when they should not. We argue that field experiment is not necessary, perfect and omnipotent, so researchers should choose appropriate research methods according to their own research objectives, research characteristics and resource constraints. We hope this paper could provide reference value for domestic researchers who plan to carry out field experiment in the future.