Project

risk preference; risk attitude; risk taking; ecological assessments; predictive modeling

Lead
Risikopräferenz wird als ein grundlegender Baustein von menschlichem Verhalten angesehen. Doch wie generalisierbar ist (durch Fragebogen oder Verhaltenstests) gemessene Risikopräferenz auf Risikoentscheidungen im Alltag?
Lay summary
Menschen unterscheiden sich stark in ihrer Risikopräferenz, was sich z.B. in unvorteilhaften Finanzentscheidungen, Glücksspiel, Substanzmissbrauch, und riskanten Entscheidungen im Freizeitbereich oder bezüglich der Gesundheit ausdrücken kann. Risikopräferenz kann deshalb den Verlauf einzelner Menschenleben substantiell beeinflussen, hat aber auch Konsequenzen für die Gesellschaft (Prä- und Intervention). Deshalb wurden verschiedene Instrumente entwickelt, um Risikopräferenz systematisch messen. Bis vor Kurzem war aber unklar, ob die verschiedenen Instrumente alle tatsächliche das gleiche Konstrukt "Risikopräferenz" messen, und wie reliabel diese Messungen über die Zeit hinweg bleiben. Mit einer umfangreichen Studie, in der 1,507 Personen 39 Messinstrumente bearbeiten haben, konnten wir substantielle Fortschritte bezüglich dieser Fragen machen. Erstaunlicherweise gibt es aber nach wie vor kaum Forschung zu den tatsächlichen Risikoentscheidungen, die Menschen im Alltag treffen – obwohl die moderne Welt mit den technologischen, ökonomischen, und gesellschaftlichen Entwicklungen ganz neue Risiken und Möglichkeiten mit sich bringt. Während wir nun also relativ gut verstehen, wie man Risikoeinstellung messen kann, wissen wir vergleichsweise wenig darüber, welche "Kriterien" im Alltag diese Messinstrumente überhaupt vorhersagen sollten. In einem ersten Teilprojekt werden wir deshalb das Risikoverhalten im modernen Leben erfassen (mit repräsentativen Studien und mobilen Assessments per Smartphone); in einem zweiten Schritt werden wir offene Fragen zur Messung von Risikopräferenz adressieren; und in einem dritten Schritt werden wir testen, wie gut das Konstrukt Risikopräferenz Verhalten im Alltag vorhersagen kann.

Direct link to Lay Summary

Last update: 23.11.2017

Active participation

Active participation

Communication	Title	Media	Place	Year

Title	Year

The construct of risk preference is considered to be a key building block of human behavior (Kahneman & Tversky, 1979; Savage, 1954). It may shape everyday decision making and is assumed to be particularly expressed in maladaptive or clinical behaviors such as impaired financial decision making, gambling, substance use, and taking excessive risks in recreational activities or regarding one’s health (Allen, Reinert, & Volk, 2001; Brodbeck, Duerrenberger, & Znoj, 2009; Frydman & Camerer, 2016; Lejuez et al., 2002; Skinner, 1982; Weber, Blais, & Betz, 2002). As individuals’ risk preferences have the potential to influence the course of entire lives, with according consequences for society (e.g., in terms of pre- or intervention; Caspi et al., 2016), the assessment of risk preference has been a focal research topic for decades. Yet, even though a plethora of different risk-taking measures has been developed (Appelt, Milch, Handgraaf, & Weber, 2011), research on the construct of risk preference has been fragmented (Schonberg, Fox, & Poldrack, 2011), leaving fundamental questions regarding the nature of the underlying construct, its measurement, and its implications for life outcomes unaddressed.We recently aimed to clarify some of these questions by adopting a large psychometric framework and collected 39 risk-taking measures from 1,507 individuals (Frey, Pedroni, Mata, Rieskamp, & Hertwig, 2017; Pedroni, Frey, Bruhin, Rieskamp, & Hertwig, 2017). Based on the solid empirical basis obtained in this study, we investigated i) the convergent validity between different measures, ii) the underlying factor structure by means of psychometric modeling, and iii) the temporal stability of risk preference as captured by these measures. Thereby, we have substantially advanced conceptual issues on the construct of risk preference and its measurement in the lab.Surprisingly, however, there is a lack of sound scientific knowledge regarding the criteria, in terms of real- life outcomes, that the different measures of risk preference ought to predict. Instead, there is a vacuum of research targeting what risks people perceive in the modern world, with its novel opportunities and threats, and the empirical evidence on the ecology of risk taking is scarce. Thus, despite of recent clarifications regarding the nature of the underlying construct (Frey et al., 2017; Highhouse, Nye, Zhang, & Rada, 2016), past studies have not satisfactorily addressed when and how the construct risk preference (as measured in the lab) generalizes to real-life outcomes.In sum, the predictive validity of the construct risk preference (and its various measures) for life outcomes remains only weakly tested. In this project, I aim to address this issue by making use of novel methodologies for ecological assessments. Specifically, the proposed studies are designed to i) create a map of risk perception and risk taking in the contemporary population, using representative methods as well as state-of-the-art methods of ecological assessment (Subproject A), ii) address a potential method confound in the typical assessment of risk preference (Subproject B), and iii) eventually assess the predictive validity of the various measures (Subproject C) by linking the first two subprojects. Ultimately, only the combination of these separate parts will permit substantial novel insights into the psychological construct of risk preference, and how we can use the established measures of risk preference for predicting real-life outcomes.