Konstantinos Katsikopoulos
MPI Berlin
Modeling violations of expected-utility theory by one-reason heuristics

People violate expected utility theory and this has been traditionally modeled by augmenting its Neo-Bernoullian (term due to Maurice Allais) weight-and-add framework, by nonlinear transformations of values and probabilities. Reinhard Selten has called this approach the "repair program" and asked for a theory of how people make decisions without using utilities and probability weights. Daniel Kahneman and Amos Tversky investigated the heuristics people use for making decisions. I analyze a model of one-reason decision-making, the priority heuristic, that does not assume that people transform values and probabilities, make tradeoffs, or use all available information. I show that the heuristic simultaneously implies, rather than just being consistent with, major violations of expected utility theory such as common consequence effects, common ratio effects, reflection effects, and the four-fold pattern of risk attitude. The intuition for these results is related to Amos Tversky's explanation for violations of transitivity in the theory of elimination-by-aspects. The results raise the possibility that loss aversion and probability weighting are not necessary for explaining people's risky choices, thus providing the basis for a new look at bounded rationality.

Marius Usher
Birkbeck College

There is recently some controversy on the issue of whether an unconscious mode of processing can enhance decision-making quality. While some results with the unconscious-deliberation paradigm (Dijksterhuis et al. 2006) supported an unconscious enhancement effect, they are subject to both replication and conceptual problems. An alternative but related position is that under some situations decision-making quality may be enhanced by the use of an intuitive/affective strategy, as compared with a sequential/analytical deliberation mode. A number of experiments are reported, consistent with this position, and indicating that intuitive thinking can lead to better results than analytic thinking under ecological decision conditions. The results are interpreted as a result of implicit value-integration by an affective decision system.

Tim Rakow
University of Essex
Helping other peoples' decisions: Graphical aids to medical risk communication

Approximately 100 members of the public participated in each of three studies examining the communication of complex medical risk information (e.g., before surgery). We assessed the efficacy of two types of graph as communication tools: survival curves and funnel plots (which doctors use to summarise outcomes, but patients rarely see). Comprehension of these graphs was generally excellent, but better when the language used to probe comprehension "matched" the labeling of the graphs (e.g., avoid labeling "survival" but discussing "mortality"). Objectively assessed and self-rated numeracy were positively related to comprehension. On the basis of these studies, we can endorse the use of these two communication tools, and can also make recommendations about how best to use them.

Dan Egan
Barclays Wealth, Behavioral Analytics
Personal Alpha -- investor satisfaction is not just about returns

Personal Alpha -- investor satisfaction is not just about returns What actually drives investor satisfaction? To my knowledge, no empirical research has been conducted on the hedonic experience of traders. As almost all modern financial theory relies on utility functions in the return space, significant hedonic value may be completely missing from most models of investor behaviour. Using data derived from a unique longitudinal investor panel drawn from the UK's largest online stockbroker, I can begin addressing this topic. From September 2008 to September 2009 over 600 active traders were surveyed every 3 months on their personal performance, the markets performance, and a host of personality factors. The extreme market movements during this period allowed us to explore the full range of potential outcomes, greatly increasing the scope of the results. Experienced returns explain only about 20% of the variance in satisfaction. Strongly beating one's benchmark (be it cash or a market index) gives a much larger hedonic effect, equivalent (in satisfaction) to a 607% annual return. Underperforming ones benchmark strongly is hedonically equivalent to making a loss of 80%. Other non-return factors include propensity to frame narrowly, financial literacy, and psychometric risk tolerance.

Gaëlle Villejoubert
Kingston University
Feeding the mind's eye: Unconscious deliberation needs visual information

Recent research suggests that when we face a choice between several options described with a large number of attributes, we make better choices if we do not consciously ponder over the alternatives but instead engage in a mindless task while our unconscious mind deliberates before finally pointing out the optimal option for us. Subsequent research attempting to replicate this "deliberation-without-attention" effect, however, provided mitigated support for its existence. The present research had two objectives. First, it aimed to improve the methodology used to test for this effect by using a choice task where attributes values were experimentally controlled. Secondly, it aimed to examine the effect of the format of presentation of the attribute values on choice quality. Participants were randomly assigned to one cell of a 2 (attribute value format) x 2 (deliberation type) design. Half of the participants received attribute values in a numerical format whereas the remaining half received them in a visual format. Within each of these conditions, half of the participants engaged in conscious deliberation before making their choice; the other half engaged unconscious deliberation. Results showed that conscious deliberation led to better decisions when the attribute values were presented in a numerical format whereas unconscious thought led to a better discrimination between products when cue values were presented in a visual format. Methodological and theoretical implications for testing the propositions of Unconscious Thought Theory will be discussed.

Patrick Brown
University of Kent
The salience of the socio-biographical for understanding decision-making: moving beyond rationality towards a "mutuality of accounts"

This paper begins by interrogating the concept of rationality and its utility in understanding or evaluating "effective" decision-making within social contexts. It is argued that analyses of individual decision-making and behaviour are often hindered by considerations of rationality (either explicit or implicit), not least because of the tautological nature of the concept. From here, a number of contexts are explored which underline the significance of multiple layers of social experience in determining the manner by which knowledge is both accessed and processed. Conversely notions of rationality would seem to encourage a fetishisation of the utility of refined or generalisable knowledge, at the expense of comprehending the intrinsic rationality in all agency and the decisive role of prior experiences which bear upon this. From this perspective it is argued that attempts (eg by professionals) to "inform" the decision-making of others are limited by the effectiveness of their communication but moreover by social structures. Rather than appealing to "rationality" they would do well to ensure a "mutuality of accounts" between scientific "evidence" and the biographical-experiential knowledge of their clients.

Dilek Önkal
Bilkent University
Effects of differential roles on group forecasts

While the behavioral research on forecasting has mostly examined the individual forecaster, organizationally-based forecasting processes tend to rely on groups to arrive at 'consensus' forecasts. Forecasting performance could also vary depending on the particular group structuring utilized in reaching a final prediction. The current study compares the forecasting performance of modified consensus groups with that of staticised groups using formal role-playing with group members assuming the roles of marketing, production and forecasting directors. It is found that, when undistorted model forecasts are given, the group forecasts (whether structured through averaging or detailed discussion of forecasts) contribute positively to forecasting accuracy. However, providing distorted initial forecasts affects accuracy with varying degrees of improvement. The results show a strong tendency to favor optimistic forecasts both for the staticised and the modified consensus group forecasts. Overall, the role modifications are found to be successful in eliciting differential adjustment behavior, effectively mimicking the disparities among different organizational roles. Current research suggests that group discussions may be an efficient method of displaying and resolving differential motivational contingencies, leading to group forecasts that perform quite well.

John Adams
University College London
Moral hazard: bonuses, seat belts and condoms

"Moral hazard" is a term that dates back to the 1600s. Until recent times its use has been mostly confined to the insurance industry to refer to behaviour that responds to changes in perceived risk.

The industry has noticed that people who have contents insurance are less careful about locking up. It has also noticed that drivers of cars with ABS brakes (superior brakes) did not have fewer accidents -- they had different accidents -- accidents consistent with high-performance cars, which is what they had become.

Why "moral" hazard? Clearly the insurance industry was disconcerted by behaviour that upset the calculations of its actuaries. Such behaviour had to be wrong -- immoral. But such behaviour is universal. Risk management is an exercise that involves striking a balance between the potential rewards and losses of decisions made in the face of uncertainty. A less judgmental term to describe this phenomenon is risk compensation. Legislators and regulators routinely ignore it.

http://john-adams.co.uk/2009/10/28/moral-hazard-bonuses-seat-belts-and-condoms/

Konstantinos Tsetos
University College London
Understanding the mechanism of perceptual and value-based decisions

Research in sequential sampling models for simple visual decisions has revealed important principles that underlie the process by which primates integrate noisy evidence across time in order to decide among a few choice alternatives. It is not clear how these principles apply in more complex situations where the presentation of the alternatives is static and decisions have to be made on the basis of subjective evaluations. Recently the sequential sampling approach was put forward to account for the intriguing pattern of preference reversals in multi-attribute decision making (i.e., similarity, compromise and attraction effects). A core assumption in these models is that despite the static nature of the stimuli, the decision maker undergoes a stochastic process of switching attention between choice aspects, integrating the momentary preference for each alternative towards a response criterion. I will present experimental data from a perceptual choice paradigm among 3 alternatives (choose the brightest among 3 spots) where the evidence is non-stationary but fluctuates, mimicking the stochastic switch of focus between dimensions that is assumed to take place in multi-attribute choice. Temporal correlations in the evidence to the 3 alternatives are manipulated in order to create situations analogous to those leading to preference reversals in value-based problems. The implications for a common selection mechanism in both perceptual and preferential decisions will be discussed with respect to the experimental data.