Petko Kusev

Exaggerated risk: prospect theory and probability weighting in risky choice

In 5 experiments, we studied precautionary decisions in which participants decided whether or not to buy insurance with specified cost against an undesirable event with specified probability and cost. We compared the risks taken for precautionary decisions with those taken for equivalent monetary gambles. Fitting these data to Tversky and Kahnemans (1992) prospect theory, we found that the weighting function required to model precautionary decisions differed from that required for monetary gambles. This result indicates a failure of the descriptive invariance axiom of expected utility theory. For precautionary decisions, people overweighted small, medium-sized, and moderately large probabilities-they exaggerated risks. This effect is not anticipated by prospect theory or experience-based decision research (Hertwig, Barron, Weber, & Erev, 2004). We found evidence that exaggerated risk is caused by the accessibility of events in memory: The weighting function varies as a function of the accessibility of events. This suggests that peoples experiences of events leak into decisions even when risk information is explicitly provided. Our findings highlight a need to investigate how variation in decision content produces variation in preferences for risk.

Judgments relative to patterns: how temporal sequence patterns affect judgments and memory

Six experiments studied relative frequency judgment and recall of sequentially presented items drawn from 2 distinct categories (i.e., city and animal). The experiments show that judged frequencies of categories of sequentially encountered stimuli are affected by certain properties of the sequence configuration. We found (a) a first-run effect whereby people overestimated the frequency of a given category when that category was the first repeated category to occur in the sequence and (b) a dissociation between judgments and recall; respondents may judge 1 event more likely than the other and yet recall more instances of the latter. Specifically, the distribution of recalled items does not correspond to the frequency estimates for the event categories, indicating that participants do not make frequency judgments by sampling their memory for individual items as implied by other accounts such as the availability heuristic (Tversky & Kahneman, 1973) and the availability process model (Hastie & Park, 1986). We interpret these findings as reflecting the operation of a judgment heuristic sensitive to sequential patterns and offer an account for the relationship between memory and judged frequencies of sequentially encountered stimuli.

Risk perceptions of cyber-security and precautionary behaviour

A quantitative empirical online study examined a set of 16 security hazards on the Internet and two comparisons in 436 UK- and US students, measuring perceptions of risk and other risk dimensions. First, perceived risk was highest for identity theft, keylogger, cyber-bullying and social engineering. Second, consistent with existing theory, significant predictors of perceived risk were voluntariness, immediacy, catastrophic potential, dread, severity of consequences and control, as well as Internet experience and frequency of Internet use. Moreover, control was a significant predictor of precautionary behaviour. Methodological implications emphasise the need for non-aggregated analysis and practical implications emphasise risk communication to Internet users.

Preferences under risk: content-dependent behavior and psychological processing

A common view in economics and psychology is that decision agents achieve their choices and express their respective preferences by computing probabilistic properties (probabilities and money) from a decision-making context (e.g., von Neumann and Morgenstern, 1947; Tversky and Kahneman, 1992; Starmer, 2000). In this computational processing, the main psychological mechanism requires that decision agents are able to integrate economic (contextual) attributes such as money and probabilities into subjective values; in other words people are able to construct and employ psycho-economic scales. Subsequently, when making a choice, decision agents are supposed to perform tradeoffs between the computed outputs (psycho-economic variables such as expected values) and certain monetary alternatives (see Kahneman and Tversky, 1979; Tversky and Kahneman, 1992; Starmer, 2000). Despite the dominance of descriptive approach to the decision-making (e.g., Kahneman and Tversky, 1979; Tversky and Kahneman, 1992), theorists (Hertwig et al., 2004; Stewart et al., 2006) have recently argued for somewhat different psychological processing in decision-making, without computations (integration of attributes) and tradeoffs. In particular, a non-utilitarian structure of preferences for risk is proposed. In this approach, decision-making is accounted for by experience with sequential events, simple binary comparisons (based on context and memory), and a threshold mechanism (Hertwig et al., 2004; Stewart et al., 2006). However, recent research (Kusev et al., 2009, 2011; Jones and Oaksford, 2011), in an effort to map the nature of human preferences, explored the role of decision-making content (the influence of memory in precautionary decision-making – Kusev et al., 2009, and transactional content on temporal and probabilistic discounting of costs – Jones and Oaksford, 2011). Specifically, we distinguish the influence of decision-making content from that of decision-making context (the description of risk); we see the content of decision-making as experiential (accumulative) cognitive storage system which represents (but not necessarily accurately) experienced events and their associate frequencies as these events occur over time. Accordingly, in this article we elaborate further on the interplay of decision-making context and content, as well as potential “decision” biases as a result of sequential experience in decision-making.

Human preferences and risky choices.

There are different views on what preferences for risks are and whether they are indicators of stable, underlying generic cognitive systems. Preferences could be conceived as an attitude toward a set of properties of context, memory, and affect – a gage of how much uncertainty one is willing to tolerate. One type of computational “descriptive” integrative decision-making theories predicts specific behavioral patterns of risky preferences. An individuals risky choice among two or more options is considered, where at least one option has an uncertain outcome1. Choices are based on the integration of probability and utility information into expected utilities, and trade-off comparisons of computed outcomes. It is assumed that there are lawful underlying patterns of risky preferences (e.g., the shapes of loss aversion and probability-weighting functions), and that these would reflect any relevant constraints in cognitive resources. In this spirit, in this research topic, Lebiere and Anderson demonstrate that their sequence-learning model, reflecting general cognitive processes in response to constraints inherent in the task environment, is superior for modeling risky choice in terms of capturing the stability that comes from previous experience. According to Luce, there are three inherently different types of people corresponding to their values of an additional utility-model parameter representing risk preference. Birnbaum demonstrates that the TAX model, in contrast to other explanations, accounts for a lack or transitivity in peoples choices. Pothos and Busemeyer show that quantum-probability theory allows the modeling of decision-making phenomena (e.g., the conjunction fallacy and violations of the sure-thing principle), which go beyond classic probability theory, because of the context- and order-dependence in quantum-probability assessment. Jones and Oaksford provide evidence for a more stable pattern of preferences in transactional decision tasks than in gambles. Given that hypothetical gambles provide results that are internally inconsistent, Baron demonstrates that a monetary-difference choice task to measure risk preference is a good indicator of peoples utility function for money. Another type of theory can be considered as “non-computational.” These theories argue for processing by establishing the role of “experience” in risky decision-making, proposing that choices are not based on the utilitarian integration of probability, and utility information, and trade-off comparisons of computed outcomes. However, yet (again) it is assumed that there are lawful underlying patterns of preferences, or people use specific processing and decision-making strategies. Stewarts results of model fitting show that, for simple risky choices, an additive (“non-integrative”) model can completely mimic a multiplicative (“integrative”) model; however, even stability of parameter values over time and across contexts in the different models does not imply correct model identification, as the parameters map onto different psychological variables. Betsch argues and provides evidence for the conceptualization of preferences as attitudes, whose stability is determined by behavior repetition and processing style. According to Hertwig and Gigerenzer, apparent inconsistencies in risky-choice behavior can be accounted for by decision-makers’ application of cognitive strategies (in particular heuristics) and the interaction of these strategies with the environment. Brandstatter contends that elicitation method strongly affects peoples choices; people use many strategies, one main candidate of which is the priority heuristic. Parducci demonstrates that range-frequency theory implies that judgments are not stable across contexts; as a result, the search for higher utility leads to reduced pleasure. Brown and Matthews show that, at least under certain conditions, rank-based models and range-based models are equivalent in that both can account for apparent range effects. Yet, still other authors explore arguments for a moderation of computational and non-computational processes of decision-making by other factors. They highlight the possibility that memory or experiences of events leak into decisions even when risk information is explicitly provided. In this research topic, Kusev and van Schaik argue and provide evidence for the idea that characteristics of (a) the decision-making context and (b) content, (c) the decision-maker (including cognitive resources and motivation), and (d) presentation format of task material (for example probability format or frequency format) all influence peoples psychological processing and subsequent risky choices. It follows then that stable behavioral patterns toward risk or the use of (single) psychological strategies do not exist. Chater, Johansson, and Hall also argue that people do not have risk preferences; rather, risky choices are shaped directly by past choices or explanations thereof. Any coherence between choices will be limited to those that share superficial features. Still other researchers provide further accounts for the apparent lack of stability of preferences. In this research topic, Fox and Tannenbaum argue that because of four specific conceptual and methodological challenges there is still a lack of evidence for stable and measurable risk preferences. Aldrovandi and van Heussen argue that the lack or degree of stability of preference in decision-making can be explained by psychological phenomena of memory; various memory phenomena lead to instability of risk preferences. Based on evidence from their neuropsychological brain research, Chen, Allen, Deb, and Humphreys argue that emotions can play a necessary functional role in decision-making, but as a consequence, emotions can alter the stability of the process. According to Dickert and Slovic, research on mental imagery and attention as underlying processes of affective responses and other research showing individual differences as moderators of these processes help explain why people do not hold stable values for saving human lives. Vlaev shows and provides evidence for the idea that trade-off inconsistency is a ubiquitous psychophysical anomaly, in which preferences between (pairs of) options are not reliable when the options are of the same qualitative type and/or differ on a single dimension. Villejoubert and Vallee-Tourangeau argue that the perspective of distributed cognition has the potential to provide a new way of conceiving of and accounting for the role of the environment in the construction of preference; the implication is that preferences may be very different when people interact with rather than respond to the environment. In conclusion, the contributions in this research topic offer a range of explanations for stability in risky choice. We are looking forward to further work that comparatively tests the validity of these different explanations and work that integrates approaches to provide a better account where this seems is appropriate.

Understanding Risky Behavior: The Influence of Cognitive, Emotional and Hormonal Factors on Decision-Making under Risk

Financial risky decisions and evaluations pervade many human everyday activities. Scientific research in such decision-making typically explores the influence of socio-economic and cognitive factors on financial behavior. However, very little research has explored the holistic influence of contextual, emotional, and hormonal factors on preferences for risk in insurance and investment behaviors. Accordingly, the goal of this review article is to address the complexity of individual risky behavior and its underlying psychological factors, as well as to critically examine current regulations on financial behavior.

Modeling judgment of sequentially presented categories using weighting and sampling without replacement

In a series of experiments, Kusev et al. (Journal of Experimental Psychology: Human Perception and Performance 37:1874–1886, 2011) studied relative-frequency judgments of items drawn from two distinct categories. The experiments showed that the judged frequencies of categories of sequentially encountered stimuli are affected by the properties of the experienced sequences. Specifically, a first-run effect was observed, whereby people overestimated the frequency of a given category when that category was the first repeated category to occur in the sequence. Here, we (1) interpret these findings as reflecting the operation of a judgment heuristic sensitive to sequential patterns, (2) present mathematical definitions of the sequences used in Kusev et al. (Journal of Experimental Psychology: Human Perception and Performance 37:1874–1886, 2011), and (3) present a mathematical formalization of the first-run effect—the judgments-relative-to-patterns model—to account for the judged frequencies of sequentially encountered stimuli. The model parameter w accounts for the effect of the length of the first run on frequency estimates, given the total sequence length. We fitted data from Kusev et al. (Journal of Experimental Psychology: Human Perception and Performance 37:1874–1886, 2011) to the model parameters, so that with increasing values of w, subsequent items in the first run have less influence on judgments. We see the role of the model as essential for advancing knowledge in the psychology of judgments, as well as in other disciplines, such as computer science, cognitive neuroscience, artificial intelligence, and human–computer interaction.

The gender pay gap : can behavioral economics provide useful insights?

People are faced with numerous decisions every day. Whether we must choose our outfit for the day, which cell phone brand to buy, what college to attend, to buy a car or house insurance, or even when or to whom to get married, decisions are a permanent presence in our daily activities. Behavioral economics is a multi-disciplinary field of study investigating how people make judgments and decisions (Camerer and Loewenstein, 2004; Heilman, 2014). Even though, from a historic point of view, behavioral economics is considered to be a relatively young field of research, the large number of studies that were undertaken and their theoretical and practical implications have made the field of behavioral economics increasingly visible among scholars. More importantly, they have also facilitated contexts to transform behavioral results into social policy programs. Starting in 2010, the UK government launched the Behavioral Insights Team, also known as The Nudge Unit, which was then followed by the Social and Behavioral Sciences Team (SBST), established by the Obama administration in 2014. Both teams aim to apply behavioral sciences, including behavioral economics, in governmental programs in order to increase peoples quality of life at lower costs. The efforts of the Nudge Unit and the SBST or other agencies and individual researchers who are trying to improve peoples overall quality of life should be supported by the research community through relevant scientific projects and by constantly finding new ways to capitalize research derived knowledge for the general use of a community.

Retrospective evaluations of sequences: Testing the predictions of a memory-based analysis

Retrospective evaluation (RE) of event sequences is known to be biased in various ways. The present paper presents a series of studies that examined the suggestion that the moments that are the most accessible in memory at the point of RE contribute to these biases. As predicted by this memory-based analysis, Experiment 1 showed that pleasantness ratings of word lists were biased by the presentation position of a negative item and by how easy the negative information was to retrieve. Experiment 2 ruled out the hypothesis that these findings were due to the dual nature of the task called upon. Experiment 3 further manipulated the memorability of the negative items--and corresponding changes in RE were as predicted. Finally, Experiment 4 extended the findings to more complex stimuli involving event narratives. Overall, the results suggest that assessments were adjusted based on the retrieval of the most readily available information.