Back to overview

Two new papers on risk preference
Posted on Wednesday, Oct 04 2017

The two core papers of the Basel-Berlin Risk Study have recently been published and are now available online (please see below for the full references). In these two publications we investigated the extent to which there is a general factor of risk preference (akin to g, the general factor of intelligence), and whether risk preference can be considered a stable psychological trait. We addressed these questions by implemented 39 risk-taking measures from three different measurement traditions: Propensity measures assessing "stated preferences", (incentivized) behavioral measures assessing "revealed preferences", and frequency measures assessing actual real-world risky activities. This battery was completed by 1,507 participants, with 109 participants completing a retest–session after a period of six months.

The results indicate that the propensity and frequency measures converge relatively well, thus forming a "positive manifold", whereas the (incentivized) behavioral measures show poor consistency (with the measures of the other measurement traditions, but also between each other). Moreover, our findings suggest that there is a broad general factor of risk preference, R, that accounts for 61% of the explained variance. This general factor is complemented by a series of domain-specific factors. Finally, the general factor proved to be highly reliable across a period of six months. To learn more about these results, please have a look at:

Frey, R., Pedroni, A., Mata, R., Rieskamp, J., & Hertwig, R. (2017). Risk preference shares the psychometric structure of major psychological traits. Science Advances, 3, e1701381. doi:10.1126/sciadv.1701381 | PDF

A second paper focuses exclusively on the behavioral measures (using the same dataset) and reports an extensive cognitive modeling analysis. The goal of this analysis was to investigate potential reasons for the lack of consistency across the various behavioral elicitation methods. People were found to differ substantially in the strategies they used in the various tasks, yet they did not do so in a very systematic way. Even at the level of model parameters, the consistency across behavioral tasks was poor, thus further calling into question the validity of behavioral measures as indicators of a person's risk preference. The detailed analyses on the behavioral tasks are published in:

Pedroni, A., Frey, R., Bruhin, A., Dutilh, G., Hertwig, R., & Rieskamp, J. (2017). The risk elicitation puzzle. Nature Human Behaviour, 1, 803-809. doi:10.1038/s41562-017-0219-x

All in all, our results suggest that risk preference has a similar psychometric structure as other major traits. In particular the observation of the general and stable factor may have important implications for future investigations of the biological foundations of risk preference. Moreover, as the lack of consistency across behavioral tasks showed, more attention needs to be given to the assessment of risk preference. For more information, please also have a look at my research section!