Current Research Projects

Expert Decision Making Under Familiar Conditions

Much of the judgement and decision-making literature has recruited convenience samples almost exclusively consisting of students. These participants have then been faced with unfamiliar tasks - ones that they have never seen before and will never come across “in the wild”. Not only might this have limited the external validity of previous results – it may also have enforced artificial constraints on what cognitive processes the participants rely on. We revisit previous judgement and decision-making studies but focus on the role that (a lack of) expertise and/or familiarity might play in mediating behaviour and cognition.

This research is co-funded by the Swedish Research School of Management and Information Technology.

To top of page

Viewing the World through a Straw

Exploring the Boundaries of Controlled Cue Integration

Research on judgment and decision making has been extensively concerned with the heuristic cognitive processes by which we negotiate the complexity of the environment to make judgments and decisions within the constraints of our bounded rationality. The “mother” of all heuristic assumptions about the environment is arguably the idea that relations between variables are linear and additive. This property of human cognition accordingly defines one of the most important constraints on our ability to understand and control complex systems, like the global economy or the environmental consequences of our actions. Somewhat surprisingly the claim for linear additive integration is mainly based on post hoc interpretations of data, and especially, on the observation that linear additive models describe judgment data well, regardless of whether the task is linear and additive or not. That people are somehow limited in this respect is clear, but exactly what the limits are is not well delineated and the causes of the constraints are not well understood. The purposes of the project are twofold: First, to systematically chart the limitations of our abilities to capture nonlinear and non-additive relations with explicit knowledge of cue-criterion relations. Second: to test and contrast the different explanations for linear additive integration that have been proposed in the literature, where one issue is if they arise from immutable cognitive limitations or reflect flexible adaptations to the proximal environment (as suggested by, e.g., Bayesian models). In this context, we advance a theory that derives this inclination from the capacity-limited and sequential nature of controlled thought.

(Supported by The Swedish Research Council.)

Publications

To top of page

Explorations of the Integration Function in Risky Decision Making

The theories of risky decision making that preoccupy the social sciences are crucial for understanding how people behave in situations of great personal and societal importance, such as choice of retirement plan and insurance policies. The field has been dominated by theories inspired by Classical Economics suggesting that people weight the outcomes by their probability into a mathematical expectation, where a rational agent decides on the option with the highest expected utility. Deviations from this notion of rationality are routinely accounted for by maintaining the integration function – probabilities and outcomes are weighted by multiplication – but postulating nonlinear responses to probability and outcome. However, no such theory is yet successful in capturing risky decision making. While multiplicative integration is often taken for granted in research on risky decision making, in other domains people are typically inclined to integrate externally provided information additively. The purpose of the project is to use psychological measurement methods to make the integration per se the subject of empirical enquiry – charting the cognitive processes that support expected value estimation, as well as the conditions where this ability collapses. We hypothesize that there is both an explicit and an implicit route by which people compute expected values, but if people are unable to recruit these processes, it dissolves into additive integration with important consequences.

(Supported by the Marcus and Amalia Foundation.)

Publications

To top of page

Understanding in an Increasingly Quantitative World

For thousands of years, as hunter-gatherers our ancestors navigated a world they thought about primarily in terms of the verbal code defined by a natural language. The brain performs complex quantitative processing in perception and motor skills, but explicit representation of quantities and mathematical laws is a new cultural artefact (indeed, most early mathematical theories, like Euclid’s Elementa, do not involve numbers, but statements about proportionality and invariance). While many cultures have a history of using money and physical measures, the last 100 years has seen an explosion in the number of quantities that lay people need to understand. Risk is expressed by probability, wealth by gross national product, memory capacity by gigabyte, energy by calories, and environmental effects by Product Environmental Footprints (PEFs), and so on. Modern societies increasingly rely on quantities to evaluate quality and progress, on levels ranging from the individual (health) to the global (environmental impact), shaping political, financial, and scientific decisions. People therefore have to navigate in an increasingly “quantitative world”. Cognitive research has addressed people’s ability to encode and process quantities and its development. There has also been recent interest in the social sciences – and in the public debate – in the management consequences of the increased use of quantitative auditing of societal activities, so called “New Public Management”. However, there is a lack of research on how this “quantitative revolution” affects the output of cognition: how we understand (represent) the world. The purpose is to investigate how people’s understanding, evaluation, and reasoning about the world is changed when they move from understanding it primarily in terms of verbal code to quantitative code; We evaluate a theory emphasizing the coding of high level beliefs at a nominal level, which thus need to interface with an increasingly quantitative world.

To top of page

Behavioral Aspects Energy-related Behaviors

This project investigates the motivations and barriers for people to adopt a more energy-efficient and environmentally-friendly behaviour. One line of research uses psychological laboratory experiments to investigate people’s cognitive ability to benefit from instant feedback on their electricity consumption by means of so called In-Home-Displays (IHDs). Another line of research investigates the mediators of so called “spill-over effects”, where policy interventions in one area lead to more (or sometimes less) environmentally-friendly behaviour in other domains. This project also connects with a number of more specific projects all involving other issues related to energy-efficiency that are performed together with Dr. Cajsa Bartusch and colleagues at the Långström laboratory that are funded by Elforsk and Energimyndigheten.

(Supported by STandUp for Energy)

Publications

To top of page

Visual Illusions and Mental Abilities

New research suggests that individual differences in perception due to differences in the involvement of recent perceptual experiences (priors) are related to personality characteristics such as hallucination proneness, creativity, openness, and autism-spectrum scores. The aim is to investigate individual differences in short-term visual priming, individual differences in seeing illusions caused by influences of priors formed during long-term development, and relation to personality characteristics. First, are the weights assigned to priors’ invariant across illusions? If so, individuals’ proneness to priming, perceptual associations, and different illusions should be related. Second, among healthy individuals, are the weights people assign to priors related to personality characteristics? If so, the results from the above mentioned perceptual tests and people’s scores on the personality tests should be related. The strength of illusions will be measured by psychophysical methods in laboratory settings, personality characteristics will be measured by standardized tests. Both correlational and experimental designs will be used. This new project bridges research in perception and personality and clinical psychology and has the potential to theory development by relating theories of perception with theories of personality characteristics.

To top of page

Decision-making Under Multiple Sources of Uncertainty

Human decision making is limited by multiple distinct sources of uncertainty, which can be grouped into two classes. First, there is a class of sources that are internal to the decision maker, such as imperfections in the quality with which our senses record information about the world and in the operations that the brain performs on this information. The second type of uncertainty is external to the decision maker and caused by ambiguities in the world. Research on decision-making under internal uncertainty has traditionally been the domain of the perceptual sciences, and decision-making under external uncertainty the domain of the cognitive sciences. However, many everyday tasks are characterized by simultaneous internal and external uncertainty, requiring the decision-maker to combine aspects of perception and cognition in a single inference. The purpose of this project is to investigate the interplay between multiple sources of uncertainty in human decision making. The insights produced by this project may have both applied and clinical implications. For example, they could ultimately be used to improve training of professionals who are required to make important decisions under multiple types of uncertainty (e.g., security personnel and radiologists who need to detect objects in x-ray images) and it could shed new light on anomalous decision-making in patients with psychiatric or neurological disorders. Moreover, the project contributes to integrating theories of perceptual decision-making with those of cognitive decision-making, which are currently developed separately and in partial conflict with each other.

(Supported by The Swedish Research Council.)

Publications

  • Stengård, E., & Van den Berg, R. (2018). Computational imperfections in visual search (under review; preprint available at bioRxiv).

To top of page

Reconceptualizing Visual Working Memory as an Adaptive System

Recent decades have seen a surge of research into characterizing the limits of visual working memory (VWM), because such knowledge is vital for understanding changes in cognitive abilities over the lifespan, clinical assessment of cognitive deficiencies, and explaining individual differences in cognitive performance. Despite great progress, there is a fierce ongoing debate about the nature of VWM. On the one hand, it is argued that VWM consists of a fixed number of “slots”, which fundamentally limits the number of items that can be kept in memory. On the other hand, it is argued that VWM is a near-continuous flexible resource that is divided across all items, which limits the quality rather than the number of items in memory. Using extensive model comparisons on data pooled together from multiple laboratories, our recent research has shown that there is strong evidence for both theories. This suggests that attempts to reject one theory in favor of the other will remain unsuccessful and it might be more constructive to examine how the two competing theories can be reconciled with each other. Our present research aims to achieve this, by breaking with the traditional idea that VWM capacity is fixed. Instead, we propose that VWM adapts both the number of items and their quality to the demands of the task that a subject is performing. If this hypothesis turns out to be correct, it will have important implications both for theories of VWM and applications that are built on these theories (such as diagnostic tests that are supposed to measure cognitive performance).

Publications

To top of page

FOLLOW UPPSALA UNIVERSITY ON

facebook
instagram
twitter
youtube
linkedin