Making Singular Risk Decisions

25 October 2019, 16:00 - 18:00

Room S2, Alison Richard Building, Cambridge, CB3 9DT

The QUALITY project at CRASSH welcomes Dr Matt Burch (University of Essex) to share some of his work on phenomenology as a method for the social sciences. 
 

Abstract:
Today’s dominant approaches to risk maintain that, whenever possible, responsible risk decisions must be based on statistical evidence and guided by the norms of some version of probability theory (Oberdiek 2017). Behind this insistence lies:

  1. Substantial empirical evidence that statistical risk prediction is superior to unaided clinical judgment (Meehl, 1954; Dawes et al. 1989)
  2. A widely popular theory that our intuitive appraisals of risk are biased by our default mental heuristics (Kahneman, Slovich, & Tversky, 1982) and implicit attitudes (Greenwald & Banaji 1995)
  3. Research suggesting that intuitive expert judgment doesn’t exist outside a narrow range of predictable environments (Kahneman & Klein 2009)

In this presentation, I will argue that probabilistic approaches to risk, though powerful in many respects, provide insufficient guidance for an important subclass of risk decisions that pervade everyday life and contemporary ‘risk work’ (Brown & Gale 2018a, 2018b), namely, singular risk decisions—one-off decisions that individuals make in the face of uncertainty with the aim of managing potential setbacks to their interests (or the interests of the person on whose behalf the decision is made). I will consider several criticisms of probabilistic approaches that highlight their limitations vis-à-vis singular risk decisions; and I will argue that taken together these criticisms suggest that we cannot make defensible singular risk decisions without appealing to moral considerations that fall outside the domain of probability theory. I will then sketch an alternative person-centred, humanistic approach to such decisions that acknowledges the strengths of probabilistic approaches while attempting to overcome some of their limitations.
 

Attendance is free but places may be limited, so please email to reserve a space.

Want to share this event? Download a poster here
 


This lecture is hosted by the ERC-funded project 'Qualitative and Quantitative Social Science: A Unified Logic of Causal Inference?'. QUALITY is funded by the European Research Council under the European Union’s Horizon 2020 Framework Programme for Research and Innovation (ERC grant agreement no. 715530)