Reliability (or “lack thereof”) of on-line preference revelation: A controlled experimental analysis

摘要

Networks permitting anonymous contributions continue to expand and flourish. In some networks, the reliability of a contribution is not of particular importance. In other settings, however, the development of a network is driven by specific purposes which make the reliability of information exchanged of significant importance. One such situation involves the use of information markets for aggregating individuals' preferences on new or emerging technologies. At this point, there remains skepticism concerning the reliability of the preference revelations in such markets and thus the resulting preference aggregations and rankings of emerging technologies. In this paper, we study the reliability of on-line preference revelation using a series of controlled laboratory experiments. Our analysis includes individuals' pre- and post-experiment rankings of technologies, individual trading and accumulation activities during an electronics market experiment, the final experimental market outcomes, and a ranking of the same technologies by a panel of experts from a Fortune 5 company. In addition, as a final step, we allowed each participant to actually select and keep a unit of one of the technologies at zero price (free). That is, we were able to observe each participant's actual final true preference from the set of technologies.