In research the general practice is for experimental studies to be derived from real-life observations, and particularly those from field research. The study here reverses this practice. A real-life situation has been discovered, the conditions of which correspond to those generally employed in risky-shift studies. The quantitative nature of the decisions made in the established group studied enables the data to be treated in comparable fashion to data derived from risk-taking and polarization studies. In the case studied here no shift in any direction was found. An examination of the possible factors accounting for this finding indicated that the two most important ones were the amount of information at the disposal of the group, and the reference system within which decisions were made, which was common to all the group members. These variables are put forward as worthy of further experimental analysis. Discussion focuses on the implications of research such as this for the external validity of experimental studies.