A re-examination of the effect of contextual group size on people's attitude to risk

Using Kahneman and Tversky's life-death decision paradigm, Wang and colleagues (e.g., Wang & Johnston, 1995; Wang, 1996a, 1996b, 1996c, 2008; Wang et al., 2001) have shown two characteristic phenomena regarding people's attitude to risk when the contextual group size is manipulated. In both positive and negative frames, people tend to take greater risks in life-death decisions as the contextual group size becomes smaller; this risk-seeking attitude is greater when framed positively than negatively. (This second characteristic often leads to the disappearance of the framing effect in small group contexts comprising of 6 or 60 people.) Their results could shed new light on the effect of contextual group size on people's risk choice. However these results are usually observed in laboratory experiments with university student samples. This study aims to examine the external validity of these results through different ways of experimentation and with a different sample base. The first characteristic was replicated in both a face-to-face interview with a randomly selected sample of the Japanese general public, and a web-based experiment with a non-student sample, but not the second.