Impulsivity is a stable, measurable, and predictive psychological trait

Impulsivity is a personality construct frequently employed to explain and predict important human behaviors. Major inconsistencies in its definition and measurement, however, have led some researchers to call for an outright rejection of impulsivity as a psychological construct. We address this highly unsatisfactory state with a large-scale, preregistered study (N = 1,676) in which each participant completed 48 measures of impulsivity derived from 10 self-report scales and 10 behavioral tasks and reported frequencies of seven impulsivity-related behaviors (e.g., impulsive buying and social media usage); a subsample (N = 196) then completed a retest session 3 mo later. We found that correlations between self-report measures were substantially higher than those between behavioral tasks and between self-report measures and behavioral tasks. Bifactor analysis of these measures exacted one general factor of impulsivity I, akin to the general intelligence factor g, and six specific factors. Factor I was related mainly to self-report measures, had high test-retest reliability, and could predict impulsivity-related behaviors better than existing measures. We further developed a scale named the adjustable impulsivity scale (AIMS) to measure I. AIMS possesses excellent psychometric properties that are largely retained in shorter versions and could predict impulsivity-related behaviors equally well as I. These findings collectively support impulsivity as a stable, measurable, and predictive trait, indicating that it may be too early to reject it as a valid and useful psychological construct. The bifactorial structure of impulsivity and AIMS, meanwhile, significantly advance the conceptualization and measurement of construct impulsivity.

Using games to understand the mind

Board, card or video games have been played by virtually every individual in the world. Games are popular because they are intuitive and fun. These distinctive qualities of games also make them ideal for studying the mind. By being intuitive, games provide a unique vantage point for understanding the inductive biases that support behaviour in more complex, ecological settings than traditional laboratory experiments. By being fun, games allow researchers to study new questions in cognition such as the meaning of 'play' and intrinsic motivation, while also supporting more extensive and diverse data collection by attracting many more participants. We describe the advantages and drawbacks of using games relative to standard laboratory-based experiments and lay out a set of recommendations on how to gain the most from using games to study cognition. We hope this Perspective will lead to a wider use of games as experimental paradigms, elevating the ecological validity, scale and robustness of research on the mind.

Boosting wisdom of the crowd for medical image annotation using training performance and task features

A crucial bottleneck in medical artificial intelligence (AI) is high-quality labeled medical datasets. In this paper, we test a large variety of wisdom of the crowd algorithms to label medical images that were initially classified by individuals recruited through an app-based platform. Individuals classified skin lesions from the International Skin Lesion Challenge 2018 into 7 different categories. There was a large dispersion in the geographical location, experience, training, and performance of the recruited individuals. We tested several wisdom of the crowd algorithms of varying complexity from a simple unweighted average to more complex Bayesian models that account for individual patterns of errors. Using a switchboard analysis, we observe that the best-performing algorithms rely on selecting top performers, weighting decisions by training accuracy, and take into account the task environment. These algorithms far exceed expert performance. We conclude by discussing the implications of these approaches for the development of medical AI.

A neural network paradigm for modeling psychometric data and estimating IRT model parameters: Cross estimation network

This paper presents a novel approach known as the cross estimation network (CEN) for fitting the datasets obtained from psychological or educational tests and estimating the parameters of item response theory (IRT) models. The CEN is comprised of two subnetworks: the person network (PN) and the item network (IN). The PN processes the response pattern of individual respondent and generates an estimate of the underlying ability, while the IN takes in the response pattern of individual item and outputs the estimates of the item parameters. Four simulation studies and an empirical study were comprehensively and rigorously conducted to investigate the performance of CEN on parameter estimation of the two-parameter logistic model under various testing scenarios. Results showed that CEN effectively fit the training data and produced accurate estimates of both person and item parameters. The trained PN and IN adhered to AI principles and acted as intelligent agents, delivering commendable evaluations for even unseen patterns of new respondents and items.

First steps into the pupillometry multiverse of developmental science

Pupillometry has been widely implemented to investigate cognitive functioning since infancy. Like most psychophysiological and behavioral measures, it implies hierarchical levels of arbitrariness in preprocessing before statistical data analysis. By means of an illustrative example, we checked the robustness of the results of a familiarization procedure that compared the impact of audiovisual and visual stimuli in 12-month-olds. We adopted a multiverse approach to pupillometry data analysis to explore the role of (1) the preprocessing phase, that is, handling of extreme values, selection of the areas of interest, management of blinks, baseline correction, participant inclusion/exclusion and (2) the modeling structure, that is, the incorporation of smoothers, fixed and random effects structure, in guiding the parameter estimation. The multiverse of analyses shows how the preprocessing steps influenced the regression results, and when visual stimuli plausibly predicted an increase of resource allocation compared with audiovisual stimuli. Importantly, smoothing time in statistical models increased the plausibility of the results compared to those nested models that do not weigh the impact of time. Finally, we share theoretical and methodological tools to move the first steps into (rather than being afraid of) the inherent uncertainty of infant pupillometry.

Modelling dataset bias in machine-learned theories of economic decision-making

Normative and descriptive models have long vied to explain and predict human risky choices, such as those between goods or gambles. A recent study reported the discovery of a new, more accurate model of human decision-making by training neural networks on a new online large-scale dataset, choices13k. Here we systematically analyse the relationships between several models and datasets using machine-learning methods and find evidence for dataset bias. Because participants' choices in stochastically dominated gambles were consistently skewed towards equipreference in the choices13k dataset, we hypothesized that this reflected increased decision noise. Indeed, a probabilistic generative model adding structured decision noise to a neural network trained on data from a laboratory study transferred best, that is, outperformed all models apart from those trained on choices13k. We conclude that a careful combination of theory and data analysis is still required to understand the complex interactions of machine-learning models and data of human risky choices.

Interactive model for predicting the oncological outcome of patients with early-stage huge hepatocellular carcinoma after hepatectomy: a multicenter population-based study

An interactive model for predicting the oncological outcome of patients with early-stage huge hepatocellular carcinoma (ES-HHCC) after hepatectomy is still lacking. This study was aimed at exploring the independent risk parameters and developing an interactive model for predicting the cancer-specific survival (CSS) of ES-HHCC. Data from patients with ES-HHCC who underwent hepatectomy were collected. The dimensionality of the clinical features was reduced by least absolute shrinkage and selection operator regression and further screened as predictors of CSS by Cox regression. Then, an interactive prediction model was developed and validated. Among the 514 screened patients, 311 and 203 of them were assigned into the training and validation cohort, respectively. Six independent variables, including alpha-fetoprotein, cirrhosis, microvascular invasion, satellite, tumor morphology, and tumor diameter, were identified and incorporated into the prediction model for CSS. The model achieved C-indices of 0.724 and 0.711 in the training and validation cohorts, respectively. Calibration curves showed general consistency in both cohorts. Compared with single predictor, the model had a better performance and greater benefit according to the time-independent receiver operating characteristic curve and decision curve analysis (P < 0.05). The calculator owned satisfactory accuracy and flexible operability for predicting the CSS of ES-HHCC, which could serve as a practical tool to stratify patients with different risks, and guide decision-making.

Automatic discovery and description of human planning strategies

Scientific discovery concerns finding patterns in data and creating insightful hypotheses that explain these patterns. Traditionally, each step of this process required human ingenuity. But the galloping development of computer chips and advances in artificial intelligence (AI) make it increasingly more feasible to automate some parts of scientific discovery. Understanding human planning is one of the fields in which AI has not yet been utilized. State-of-the-art methods for discovering new planning strategies still rely on manual data analysis. Data about the process of human planning is often used to group similar behaviors together. Researchers then use this data to formulate verbal descriptions of the strategies which might underlie those groups of behaviors. In this work, we leverage AI to automate these two steps of scientific discovery. We introduce a method for automatic discovery and description of human planning strategies from process-tracing data collected with the Mouselab-MDP paradigm. Our method utilizes a new algorithm, called Human-Interpret, that performs imitation learning to describe sequences of planning operations in terms of a procedural formula and then translates that formula to natural language. We test our method on a benchmark data set that researchers have previously scrutinized manually. We find that the descriptions of human planning strategies that we obtain automatically are about as understandable as human-generated descriptions. They also cover a substantial proportion of relevant types of human planning strategies that had been discovered manually. Our method saves scientists' time and effort, as all the reasoning about human planning is done automatically. This might make it feasible to more rapidly scale up the search for yet undiscovered cognitive strategies that people use for planning and decision-making to many new decision environments, populations, tasks, and domains. Given these results, we believe that the presented work may accelerate scientific discovery in psychology, and due to its generality, extend to problems from other fields.

Artificial intelligence and illusions of understanding in scientific research

Scientists are enthusiastically imagining ways in which artificial intelligence (AI) tools might improve research. Why are AI tools so attractive and what are the risks of implementing them across the research pipeline? Here we develop a taxonomy of scientists' visions for AI, observing that their appeal comes from promises to improve productivity and objectivity by overcoming human shortcomings. But proposed AI solutions can also exploit our cognitive limitations, making us vulnerable to illusions of understanding in which we believe we understand more about the world than we actually do. Such illusions obscure the scientific community's ability to see the formation of scientific monocultures, in which some types of methods, questions and viewpoints come to dominate alternative approaches, making science less innovative and more vulnerable to errors. The proliferation of AI tools in science risks introducing a phase of scientific enquiry in which we produce more but understand less. By analysing the appeal of these tools, we provide a framework for advancing discussions of responsible knowledge production in the age of AI.

Machine Learning Prediction of Molecular Binding Profiles on Metal-Porphyrin via Spectroscopic Descriptors

The study of molecular adsorption is crucial for understanding various chemical processes. Spectroscopy offers a convenient and non-invasive way of probing structures of adsorbed states and can be used for real-time observation of molecular binding profiles, including both structural and energetic information. However, deciphering atomic structures from spectral information using the first-principles approach is computationally expensive and time-consuming because of the sophistication of recording spectra, chemical structures, and their relationship. Here, we demonstrate the feasibility of a data-driven machine learning approach for predicting binding energy and structural information directly from vibrational spectra of the adsorbate by using CO adsorption on iron porphyrin as an example. Our trained machine learning model is not only interpretable but also readily transferred to similar metal-nitrogen-carbon systems with comparable accuracy. This work shows the potential of using structure-encoded spectroscopic descriptors in machine learning models for the study of adsorbed states of molecules on transition metal complexes.

Designing optimal behavioral experiments using machine learning

Computational models are powerful tools for understanding human cognition and behavior. They let us express our theories clearly and precisely and offer predictions that can be subtle and often counter-intuitive. However, this same richness and ability to surprise means our scientific intuitions and traditional tools are ill-suited to designing experiments to test and compare these models. To avoid these pitfalls and realize the full potential of computational modeling, we require tools to design experiments that provide clear answers about what models explain human behavior and the auxiliary assumptions those models must make. Bayesian optimal experimental design (BOED) formalizes the search for optimal experimental designs by identifying experiments that are expected to yield informative data. In this work, we provide a tutorial on leveraging recent advances in BOED and machine learning to find optimal experiments for any kind of model that we can simulate data from, and show how by-products of this procedure allow for quick and straightforward evaluation of models and their parameters against real experimental data. As a case study, we consider theories of how people balance exploration and exploitation in multi-armed bandit decision-making tasks. We validate the presented approach using simulations and a real-world experiment. As compared to experimental designs commonly used in the literature, we show that our optimal designs more efficiently determine which of a set of models best account for individual human behavior, and more efficiently characterize behavior given a preferred model. At the same time, formalizing a scientific question such that it can be adequately addressed with BOED can be challenging and we discuss several potential caveats and pitfalls that practitioners should be aware of. We provide code to replicate all analyses as well as tutorial notebooks and pointers to adapt the methodology to different experimental settings.

Multiple timescales of temporal context in risky choice: Behavioral identification and relationships to physiological arousal

Context-dependence is fundamental to risky monetary decision-making. A growing body of evidence suggests that temporal context, or recent events, alters risk-taking at a minimum of three timescales: immediate (e.g. trial-by-trial), neighborhood (e.g. a group of consecutive trials), and global (e.g. task-level). To examine context effects, we created a novel monetary choice set with intentional temporal structure in which option values shifted between multiple levels of value magnitude ("contexts") several times over the course of the task. This structure allowed us to examine whether effects of each timescale were simultaneously present in risky choice behavior and the potential mechanistic role of arousal, an established correlate of risk-taking, in context-dependency. We found that risk-taking was sensitive to immediate, neighborhood, and global timescales: risk-taking decreased following large (vs. small) outcome amounts, increased following large positive (but not negative) shifts in context, and increased when cumulative earnings exceeded expectations. We quantified arousal with skin conductance responses, which were related to the global timescale, increasing with cumulative earnings, suggesting that physiological arousal captures a task-level assessment of performance. Our results both replicate and extend prior research by demonstrating that risky decision-making is consistently dynamic at multiple timescales and that the role of arousal in risk-taking extends to some, but not all timescales of context-dependence.

A multi-stage anticipated surprise model with dynamic expectation for economic decision-making

There are many modeling works that aim to explain people's behaviors that violate classical economic theories. However, these models often do not take into full account the multi-stage nature of real-life problems and people's tendency in solving complicated problems sequentially. In this work, we propose a descriptive decision-making model for multi-stage problems with perceived post-decision information. In the model, decisions are chosen based on an entity which we call the 'anticipated surprise'. The reference point is determined by the expected value of the possible outcomes, which we assume to be dynamically changing during the mental simulation of a sequence of events. We illustrate how our formalism can help us understand prominent economic paradoxes and gambling behaviors that involve multi-stage or sequential planning. We also discuss how neuroscience findings, like prediction error signals and introspective neuronal replay, as well as psychological theories like affective forecasting, are related to the features in our model. This provides hints for future experiments to investigate the role of these entities in decision-making.

Harnessing the flexibility of neural networks to predict dynamic theoretical parameters underlying human choice behavior

Reinforcement learning (RL) models are used extensively to study human behavior. These rely on normative models of behavior and stress interpretability over predictive capabilities. More recently, neural network models have emerged as a descriptive modeling paradigm that is capable of high predictive power yet with limited interpretability. Here, we seek to augment the expressiveness of theoretical RL models with the high flexibility and predictive power of neural networks. We introduce a novel framework, which we term theoretical-RNN (t-RNN), whereby a recurrent neural network is trained to predict trial-by-trial behavior and to infer theoretical RL parameters using artificial data of RL agents performing a two-armed bandit task. In three studies, we then examined the use of our approach to dynamically predict unseen behavior along with time-varying theoretical RL parameters. We first validate our approach using synthetic data with known RL parameters. Next, as a proof-of-concept, we applied our framework to two independent datasets of humans performing the same task. In the first dataset, we describe differences in theoretical RL parameters dynamic among clinical psychiatric vs. healthy controls. In the second dataset, we show that the exploration strategies of humans varied dynamically in response to task phase and difficulty. For all analyses, we found better performance in the prediction of actions for t-RNN compared to the stationary maximum-likelihood RL method. We discuss the use of neural networks to facilitate the estimation of latent RL parameters underlying choice behavior.

The Future of Decisions From Experience: Connecting Real-World Decision Problems to Cognitive Processes

In many important real-world decision domains, such as finance, the environment, and health, behavior is strongly influenced by experience. Renewed interest in studying this influence led to important advancements in the understanding of these decisions from experience (DfE) in the last 20 years. Building on this literature, we suggest ways the standard experimental design should be extended to better approach important real-world DfE. These extensions include, for example, introducing more complex choice situations, delaying feedback, and including social interactions. When acting upon experiences in these richer and more complicated environments, extensive cognitive processes go into making a decision. Therefore, we argue for integrating cognitive processes more explicitly into experimental research in DfE. These cognitive processes include attention to and perception of numeric and nonnumeric experiences, the influence of episodic and semantic memory, and the mental models involved in learning processes. Understanding these basic cognitive processes can advance the modeling, understanding and prediction of DfE in the laboratory and in the real world. We highlight the potential of experimental research in DfE for theory integration across the behavioral, decision, and cognitive sciences. Furthermore, this research could lead to new methodology that better informs decision-making and policy interventions.

Developmental psychologists should adopt citizen science to improve generalization and reproducibility

Widespread failures of replication and generalization are, ironically, a scientific triumph, in that they confirm the fundamental metascientific theory that underlies our field. Generalizable and replicable findings require testing large numbers of subjects from a wide range of demographics with a large, randomly-sampled stimulus set, and using a variety of experimental parameters. Because few studies accomplish any of this, meta-scientists predict that findings will frequently fail to replicate or generalize. We argue that to be more robust and replicable, developmental psychology needs to find a mechanism for collecting data at greater scale and from more diverse populations. Luckily, this mechanism already exists: Citizen science, in which large numbers of uncompensated volunteers provide data. While best-known for its contributions to astronomy and ecology, citizen science has also produced major findings in neuroscience and psychology, and increasingly in developmental psychology. We provide examples, address practical challenges, discuss limitations, and compare to other methods of obtaining large datasets. Ultimately, we argue that the range of studies where it makes sense *not* to use citizen science is steadily dwindling.

Statistical prediction alone cannot identify good models of behavior

The dissociation between statistical prediction and scientific explanation advanced by Bowers et al. for studies of vision using deep neural networks is also observed in several other domains of behavior research, and is in fact unavoidable when fitting large models such as deep nets and other supervised learners, with weak theoretical commitments, to restricted samples of highly stochastic behavioral phenomena.

Risky decisions are influenced by individual attributes as a function of risk preference

It has long been assumed in economic theory that multi-attribute decisions involving several attributes or dimensions - such as probabilities and amounts of money to be earned during risky choices - are resolved by first combining the attributes of each option to form an overall expected value and then comparing the expected values of the alternative options, using a unique evidence accumulation process. A plausible alternative would be performing independent comparisons between the individual attributes and then integrating the results of the comparisons afterwards. Here, we devise a novel method to disambiguate between these types of models, by orthogonally manipulating the expected value of choice options and the relative salience of their attributes. Our results, based on behavioral measures and drift-diffusion models, provide evidence in favor of the framework where information about individual attributes independently impacts deliberation. This suggests that risky decisions are resolved by running in parallel multiple comparisons between the separate attributes - possibly alongside an additional comparison of expected value. This result stands in contrast with the assumption of standard economic theory that choices require a unique comparison of expected values and suggests that at the cognitive level, decision processes might be more distributed than commonly assumed. Beyond our planned analyses, we also discovered that attribute salience affects people of different risk preference type in different ways: risk-averse participants seem to focus more on probability, except when monetary amount is particularly high; risk-neutral/seeking participants, in contrast, seem to focus more on monetary amount, except when probability is particularly low.

Using deep neural networks as a guide for modeling human planning

When developing models in cognitive science, researchers typically start with their own intuitions about human behavior in a given task and then build in mechanisms that explain additional aspects of the data. This refinement step is often hindered by how difficult it is to distinguish the unpredictable randomness of people's decisions from meaningful deviations between those decisions and the model. One solution for this problem is to compare the model against deep neural networks trained on behavioral data, which can detect almost any pattern given sufficient data. Here, we apply this method to the domain of planning with a heuristic search model for human play in 4-in-a-row, a combinatorial game where participants think multiple steps into the future. Using a data set consisting of 10,874,547 games, we train deep neural networks to predict human moves and find that they accurately do so while capturing meaningful patterns in the data. Thus, deviations between the model and the best network allow us to identify opportunities for model improvement despite starting with a model that has undergone substantial testing in previous work. Based on this analysis, we add three extensions to the model that range from a simple opening bias to specific adjustments regarding endgame planning. Overall, our work demonstrates the advantages of model comparison with a high-performance deep neural network as well as the feasibility of scaling cognitive models to massive data sets for systematically investigating the processes underlying human sequential decision-making.

Adversarial Dynamics in Centralized Versus Decentralized Intelligent Systems

Artificial intelligence (AI) is often used to predict human behavior, thus potentially posing limitations to individuals' and collectives' freedom to act. AI's most controversial and contested applications range from targeted advertisements to crime prevention, including the suppression of civil disorder. Scholars and civil society watchdogs are discussing the oppressive dangers of AI being used by centralized institutions, like governments or private corporations. Some suggest that AI gives asymmetrical power to governments, compared to their citizens. On the other hand, civil protests often rely on distributed networks of activists without centralized leadership or planning. Civil protests create an adversarial tension between centralized and decentralized intelligence, opening the question of how distributed human networks can collectively adapt and outperform a hostile centralized AI trying to anticipate and control their activities. This paper leverages multi-agent reinforcement learning to simulate dynamics within a human-machine hybrid society. We ask how decentralized intelligent agents can collectively adapt when competing with a centralized predictive algorithm, wherein prediction involves suppressing coordination. In particular, we investigate an adversarial game between a collective of individual learners and a central predictive algorithm, each trained through deep Q-learning. We compare different predictive architectures and showcase conditions in which the adversarial nature of this dynamic pushes each intelligence to increase its behavioral complexity to outperform its counterpart. We further show that a shared predictive algorithm drives decentralized agents to align their behavior. This work sheds light on the totalitarian danger posed by AI and provides evidence that decentrally organized humans can overcome its risks by developing increasingly complex coordination strategies.

Semantic determinants of memorability

We examine why some words are more memorable than others by using predictive machine learning models applied to word recognition and recall datasets. Our approach provides more accurate out-of-sample predictions for recognition and recall than previous psychological models, and outperforms human participants in new studies of memorability prediction. Our approach's predictive power stems from its ability to capture the semantic determinants of memorability in a data-driven manner. We identify which semantic categories are important for memorability and show that, unlike features such as word frequency that influence recognition and recall differently, the memorability of semantic categories is consistent across recognition and recall. Our paper sheds light on the complex psychological drivers of memorability, and in doing so illustrates the power of machine learning methods for psychological theory development.

Computational and systems neuroscience: The next 20 years

Over the past 20 years, neuroscience has been propelled forward by theory-driven experimentation. We consider the future outlook for the field in the age of big neural data and powerful artificial intelligence models.

Causal inference in cognitive neuroscience

Causal inference is a key step in many research endeavors in cognitive science and neuroscience, and particularly cognitive neuroscience. Statistical knowledge is sufficient for prediction and diagnosis, but causal knowledge is required for action and intervention. Most statistics courses and textbooks emphasize the difficulty of causal inference, focusing on the maxim that "correlation does not mean causation": there can be multiple causal possibilities, often many of them, consistent with given observed statistics. This paper focuses instead on the conceptual issues and assumptions that confront causal and other kinds of inference, primarily focusing on cognitive neuroscience. We connect inference methods with goals and challenges, and provide concrete guidance about how to select appropriate tools for the scientific task. This article is categorized under: Psychology > Theory and Methods Philosophy > Foundations of Cognitive Science.

Carving joints into nature: reengineering scientific concepts in light of concept-laden evidence

A new wave of proposals suggests that scientists must reassess scientific concepts in light of accumulated evidence. However, reengineering scientific concepts in light of data is challenging because scientific concepts affect the evidence itself in multiple ways. Among other possible influences, concepts (i) prime scientists to overemphasize within-concept similarities and between-concept differences; (ii) lead scientists to measure conceptually relevant dimensions more accurately; (iii) serve as units of scientific experimentation, communication, and theory-building; and (iv) affect the phenomena themselves. When looking for improved ways to carve nature at its joints, scholars must take the concept-laden nature of evidence into account to avoid entering a vicious circle of concept-evidence mutual substantiation.

AI and the transformation of social science research

Careful bias management and data fidelity are key.

A quasi-comprehensive exploration of the mechanisms of spatial working memory

Why are some spatial patterns remembered more easily than others? There are many possible mechanisms underlying spatial working memory function. Here, the author explores different mechanisms simultaneously in a single conceptual model. He conducts a large-scale experiment (35.4 million responses used to measure human observers' spatial working memory across 80,000 patterns) and builds a convolutional neural network as a benchmark for what is expected to be explainable. The author then creates a quasi-comprehensive exploration model of spatial working memory based on classic concepts, as well as new notions, including spatial uncertainty, Bayesian integration, out-of-range responses, averaging, grouping, categorical memory, line detection, gap detection, blurring, lateral inhibition, chunking, multiple spatial-frequency channels, redundancy, response bias and random guess. This model provides a tentative overarching framework for the mechanisms of spatial working memory.

Modelling human behaviour in cognitive tasks with latent dynamical systems

Response time data collected from cognitive tasks are a cornerstone of psychology and neuroscience research, yet existing models of these data either make strong assumptions about the data-generating process or are limited to modelling single trials. We introduce task-DyVA, a deep learning framework in which expressive dynamical systems are trained to reproduce sequences of response times observed in data from individual human subjects. Models fitted to a large task-switching dataset captured subject-specific behavioural differences with high temporal precision, including task-switching costs. Through perturbation experiments and analyses of the models' latent dynamics, we find support for a rational account of switch costs in terms of a stability-flexibility trade-off. Thus, our framework can be used to discover interpretable cognitive theories that explain how the brain dynamically gives rise to behaviour.

Computational Scientific Discovery in Psychology

Scientific discovery is a driving force for progress involving creative problem-solving processes to further our understanding of the world. The process of scientific discovery has historically been intensive and time-consuming; however, advances in computational power and algorithms have provided an efficient route to make new discoveries. Complex tools using artificial intelligence (AI) can efficiently analyze data as well as generate new hypotheses and theories. Along with AI becoming increasingly prevalent in our daily lives and the services we access, its application to different scientific domains is becoming more widespread. For example, AI has been used for the early detection of medical conditions, identifying treatments and vaccines (e.g., against COVID-19), and predicting protein structure. The application of AI in psychological science has started to become popular. AI can assist in new discoveries both as a tool that allows more freedom to scientists to generate new theories and by making creative discoveries autonomously. Conversely, psychological concepts such as heuristics have refined and improved artificial systems. With such powerful systems, however, there are key ethical and practical issues to consider. This article addresses the current and future directions of computational scientific discovery generally and its applications in psychological science more specifically.

Degrees of algorithmic equivalence between the brain and its DNN models

Deep neural networks (DNNs) have become powerful and increasingly ubiquitous tools to model human cognition, and often produce similar behaviors. For example, with their hierarchical, brain-inspired organization of computations, DNNs apparently categorize real-world images in the same way as humans do. Does this imply that their categorization algorithms are also similar? We have framed the question with three embedded degrees that progressively constrain algorithmic similarity evaluations: equivalence of (i) behavioral/brain responses, which is current practice, (ii) the stimulus features that are processed to produce these outcomes, which is more constraining, and (iii) the algorithms that process these shared features, the ultimate goal. To improve DNNs as models of cognition, we develop for each degree an increasingly constrained benchmark that specifies the epistemological conditions for the considered equivalence.

Mapping of machine learning approaches for description, prediction, and causal inference in the social and health sciences

Machine learning (ML) methodology used in the social and health sciences needs to fit the intended research purposes of description, prediction, or causal inference. This paper provides a comprehensive, systematic meta-mapping of research questions in the social and health sciences to appropriate ML approaches by incorporating the necessary requirements to statistical analysis in these disciplines. We map the established classification into description, prediction, counterfactual prediction, and causal structural learning to common research goals, such as estimating prevalence of adverse social or health outcomes, predicting the risk of an event, and identifying risk factors or causes of adverse outcomes, and explain common ML performance metrics. Such mapping may help to fully exploit the benefits of ML while considering domain-specific aspects relevant to the social and health sciences and hopefully contribute to the acceleration of the uptake of ML applications to advance both basic and applied social and health sciences research.

Cognition-Enhanced Machine Learning for Better Predictions with Limited Data

The fields of machine learning (ML) and cognitive science have developed complementary approaches to computationally modeling human behavior. ML's primary concern is maximizing prediction accuracy; cognitive science's primary concern is explaining the underlying mechanisms. Cross-talk between these disciplines is limited, likely because the tasks and goals usually differ. The domain of e-learning and knowledge acquisition constitutes a fruitful intersection for the two fields' methodologies to be integrated because accurately tracking learning and forgetting over time and predicting future performance based on learning histories are central to developing effective, personalized learning tools. Here, we show how a state-of-the-art ML model can be enhanced by incorporating insights from a cognitive model of human memory. This was done by exploiting the predictive performance equation's (PPE) narrow but highly specialized domain knowledge with regard to the temporal dynamics of learning and forgetting. Specifically, the PPE was used to engineer timing-related input features for a gradient-boosted decision trees (GBDT) model. The resulting PPE-enhanced GBDT outperformed the default GBDT, especially under conditions in which limited data were available for training. Results suggest that integrating cognitive and ML models could be particularly productive if the available data are too high-dimensional to be explained by a cognitive model but not sufficiently large to effectively train a modern ML algorithm. Here, the cognitive model's insights pertaining to only one aspect of the data were enough to jump-start the ML model's ability to make predictions-a finding that holds promise for future explorations.

Interdependent Autonomous Human-Machine Systems: The Complementarity of Fitness, Vulnerability and Evolution

For the science of autonomous human-machine systems, traditional causal-time interpretations of reality in known contexts are sufficient for rational decisions and actions to be taken, but not for uncertain or dynamic contexts, nor for building the best teams. First, unlike game theory where the contexts are constructed for players, or machine learning where contexts must be stable, when facing uncertainty or conflict, a rational process is insufficient for decisions or actions to be taken; second, as supported by the literature, rational explanations cannot disaggregate human-machine teams. In the first case, interdependent humans facing uncertainty spontaneously engage in debate over complementary tradeoffs in a search for the best path forward, characterized by maximum entropy production (MEP); however, in the second case, signified by a reduction in structural entropy production (SEP), interdependent team structures make it rationally impossible to discern what creates better teams. In our review of evidence for SEP-MEP complementarity for teams, we found that structural redundancy for top global oil producers, replicated for top global militaries, impedes interdependence and promotes corruption. Next, using UN data for Middle Eastern North African nations plus Israel, we found that a nation's structure of education is significantly associated with MEP by the number of patents it produces; this conflicts with our earlier finding that a U.S. Air Force education in air combat maneuvering was not associated with the best performance in air combat, but air combat flight training was. These last two results exemplify that SEP-MEP interactions by the team's best members are made by orthogonal contributions. We extend our theory to find that competition between teams hinges on vulnerability, a complementary excess of SEP and reduced MEP, which generalizes to autonomous human-machine systems.

Quantitatively Determining Surface-Adsorbate Properties from Vibrational Spectroscopy with Interpretable Machine Learning

Learning microscopic properties of a material from its macroscopic measurables is a grand and challenging goal in physical science. Conventional wisdom is to first identify material structures exploiting characterization tools, such as spectroscopy, and then to infer properties of interest, often with assistance of theory and simulations. This indirect approach has limitations due to the accumulation of errors from retrieving structures from spectral signals and the lack of quantitative structure-property relationship. A new pathway directly from spectral signals to microscopic properties is highly desirable, as it would offer valuable guidance toward materials evaluation and design via spectroscopic measurements. Herein, we exploit machine-learned vibrational spectroscopy to establish quantitative spectrum-property relationships. Key interaction properties of substrate-adsorbate systems, including adsorption energy and charge transfer, are quantitatively determined directly from Infrared and Raman spectroscopic signals of the adsorbates. The machine-learned spectrum-property relationships are presented as mathematical formulas, which are physically interpretable and therefore transferrable to a series of metal/alloy surfaces. The demonstrated ability of quantitative determination of hard-to-measure microscopic properties using machine-learned spectroscopy will significantly broaden the applicability of conventional spectroscopic techniques for materials design and high throughput screening under operando conditions.

Sharing Rewards Undermines Coordinated Hunting

Coordinated hunting is widely observed in animals, and sharing rewards is often considered a major incentive for its success. While current theories about the role played by sharing in coordinated hunting are based on correlational evidence, we reveal the causal roles of sharing rewards through computational modeling with a state-of-the-art Multi-agent Reinforcement Learning (MARL) algorithm. We show that counterintuitively, while selfish agents reach robust coordination, sharing rewards undermines coordination. Hunting coordination modeled through sharing rewards (1) suffers from the free-rider problem, (2) plateaus at a small group size, and (3) is not a Nash equilibrium. Moreover, individually rewarded predators outperform predators that share rewards, especially when the hunting is difficult, the group size is large, and the action cost is high. Our results shed new light on the actual importance of prosocial motives for successful coordination in nonhuman animals and suggest that sharing rewards might simply be a byproduct of hunting, instead of a design strategy aimed at facilitating group coordination. This also highlights that current artificial intelligence modeling of human-like coordination in a group setting that assumes rewards sharing as a motivator (e.g., MARL) might not be adequately capturing what is truly necessary for successful coordination.

What Can Game Theory Tell Us about an AI ‘Theory of Mind’?

Game theory includes a rich source of methods for analysing strategic interactions where there are a small number of agents, each having only a few choices. In more complex settings though, where there are many choices over indefinite time horizons involving large social groups, these methods are unlikely to fully capture the causes of agent behaviour. If agents are able to simplify the task of understanding what others might do by modelling the constraints of others, particularly unobservable cognitive constraints, then the possible behavioural outcomes can be similarly restricted, thereby reducing the complexity of a social interaction. Having a cognitive representation of the unobserved causal states of others is an aspect of a ‘Theory of Mind’ and it plays a central role in the psychology of social interactions. In this article I examine a selection of results on the theory of mind and connect these with the ‘game theory of mind’ to draw conclusions regarding the complexity of one-on-one and large-scale social coordination. To make this explicit, I will illustrate the relationship between the two psychological terms ‘introspection’ and ‘theory of mind’ and the economic analysis of game theory, while retaining as much as possible of the richness of the psychological concepts. It will be shown that game theory plays an important role in modelling interpersonal relationships for both biological and artificial agents, but it is not yet the whole story, and some psychological refinements to game theory are discussed.

Machine learning approaches for biomolecular, biophysical, and biomaterials research

A fluent conversation with a virtual assistant, person-tailored news feeds, and deep-fake images created within seconds-all those things that have been unthinkable for a long time are now a part of our everyday lives. What these examples have in common is that they are realized by different means of machine learning (ML), a technology that has fundamentally changed many aspects of the modern world. The possibility to process enormous amount of data in multi-hierarchical, digital constructs has paved the way not only for creating intelligent systems but also for obtaining surprising new insight into many scientific problems. However, in the different areas of biosciences, which typically rely heavily on the collection of time-consuming experimental data, applying ML methods is a bit more challenging: Here, difficulties can arise from small datasets and the inherent, broad variability, and complexity associated with studying biological objects and phenomena. In this Review, we give an overview of commonly used ML algorithms (which are often referred to as "machines") and learning strategies as well as their applications in different bio-disciplines such as molecular biology, drug development, biophysics, and biomaterials science. We highlight how selected research questions from those fields were successfully translated into machine readable formats, discuss typical problems that can arise in this context, and provide an overview of how to resolve those encountered difficulties.

Individual and collective learning in groups facing danger

While social networks jeopardize people’s well-being by working as diffusion pathways of falsehood, they may also help people overcome the challenge of misinformation with time and experience. Here I examine how social networks provide learning facilitation using an experiment involving an iterated decision-making game simulating an unpredictable situation faced by a group (2786 subjects in 120 groups). This study shows that, while social networks initially spread false information and suppress necessary actions, with tie rewiring, on the other hand, they facilitate improvement in people's decision-making across time. It also shows that the network's learning facilitation results from the integration of individual experiences into structural changes. In sum, social networks can support collective learning when they are built through people's experiences and accumulated relationships.

Risk Determination versus Risk Perception: A New Model of Reality for Human–Machine Autonomy

We review the progress in developing a science of interdependence applied to the determinations and perceptions of risk for autonomous human–machine systems based on a case study of the Department of Defense’s (DoD) faulty determination of risk in a drone strike in Afghanistan; the DoD’s assessment was rushed, suppressing alternative risk perceptions. We begin by contrasting the lack of success found in a case study from the commercial sphere (Facebook’s use of machine intelligence to find and categorize “hate speech”). Then, after the DoD case study, we draw a comparison with the Department of Energy’s (DOE) mismanagement of its military nuclear wastes that created health risks to the public, DOE employees, and the environment. The DOE recovered by defending its risk determinations and challenging risk perceptions in public. We apply this process to autonomous human–machine systems. The result from this review is a major discovery about the costly suppression of risk perceptions to best determine actual risks, whether for the military, business, or politics. For autonomous systems, we conclude that the determinations of actual risks need to be limited in scope as much as feasible; and that a process of free and open debate needs to be adopted that challenges the risk perceptions arising in situations facing uncertainty as the best, and possibly the only, path forward to a solution.

Utility of Machine Learning Approach with Neuropsychological Tests in Predicting Functional Impairment of Alzheimer's Disease

Background:

In assessing the levels of clinical impairment in dementia, a summary index of neuropsychological batteries has been widely used in describing the overall functional status.

Objective:

It remains unexamined how complex patterns of the test performances can be utilized to have specific predictive meaning when the machine learning approach is applied.

Methods:

In this study, the neuropsychological battery (CERAD-K) and assessment of functioning level (Clinical Dementia Rating scale and Instrumental Activities of Daily Living) were administered to 2,642 older adults with no impairment (n = 285), mild cognitive impairment (n = 1,057), and Alzheimer’s disease (n = 1,300). Predictive accuracy on functional impairment level with the linear models of the single total score or multiple subtest scores (Model 1, 2) and support vector regression with low or high complexity (Model 3, 4) were compared across different sample sizes.

Results:

The linear models (Model 1, 2) showed superior performance with relatively smaller sample size, while nonlinear models with low and high complexity (Model 3, 4) showed an improved accuracy with a larger dataset. Unlike linear models, the nonlinear models showed a gradual increase in the predictive accuracy with a larger sample size (n > 500), especially when the model training is allowed to exploit complex patterns of the dataset.

Conclusion:

Our finding suggests that nonlinear models can predict levels of functional impairment with a sufficient dataset. The summary index of the neuropsychological battery can be augmented for specific purposes, especially in estimating the functional status of dementia.

Magnitude-sensitivity: rethinking decision-making

Magnitude-sensitivity refers to the result that performance in decision-making, across domains and organisms, is affected by the total value of the possible alternatives. This simple result offers a window into fundamental issues in decision-making and has led to a reconsideration of ecological decision-making, prominent computational models of decision-making, and optimal decision-making. Moreover, magnitude-sensitivity has inspired the design of new robotic systems that exploit natural solutions and apply optimal decision-making policies. In this article, we review the key theoretical and empirical results about magnitude-sensitivity and highlight the importance that this phenomenon has for the understanding of decision-making. Furthermore, we discuss open questions and ideas for future research.

Testing models at the neural level reveals how the brain computes subjective value

In recent years, models have played an increasingly important role for understanding the brain in cognitive, behavioral, and systems neuroscience. Decision neuroscience in particular has benefitted greatly from the application of economic models of choice preferences to neural data. However, an often-overlooked aspect is that many models of preferences have a generic problem—they make extremely similar behavioral predictions. Here, we demonstrate that to understand the mechanisms of valuation in the brain, it is useful to compare models of choice preferences not only at the behavioral but also at the neural level.

Decisions are based on the subjective values of choice options. However, subjective value is a theoretical construct and not directly observable. Strikingly, distinct theoretical models competing to explain how subjective values are assigned to choice options often make very similar behavioral predictions, which poses a major difficulty for establishing a mechanistic, biologically plausible explanation of decision-making based on behavior alone. Here, we demonstrate that model comparison at the neural level provides insights into model implementation during subjective value computation even though the distinct models parametrically identify common brain regions as computing subjective value. We show that frontal cortical regions implement a model based on the statistical distributions of available rewards, whereas intraparietal cortex and striatum compute subjective value signals according to a model based on distortions in the representations of probabilities. Thus, better mechanistic understanding of how cognitive processes are implemented arises from model comparisons at the neural level, over and above the traditional approach of comparing models at the behavioral level alone.

The Transporter-Mediated Cellular Uptake and Efflux of Pharmaceutical Drugs and Biotechnology Products: How and Why Phospholipid Bilayer Transport Is Negligible in Real Biomembranes

Over the years, my colleagues and I have come to realise that the likelihood of pharmaceutical drugs being able to diffuse through whatever unhindered phospholipid bilayer may exist in intact biological membranes in vivo is vanishingly low. This is because (i) most real biomembranes are mostly protein, not lipid, (ii) unlike purely lipid bilayers that can form transient aqueous channels, the high concentrations of proteins serve to stop such activity, (iii) natural evolution long ago selected against transport methods that just let any undesirable products enter a cell, (iv) transporters have now been identified for all kinds of molecules (even water) that were once thought not to require them, (v) many experiments show a massive variation in the uptake of drugs between different cells, tissues, and organisms, that cannot be explained if lipid bilayer transport is significant or if efflux were the only differentiator, and (vi) many experiments that manipulate the expression level of individual transporters as an independent variable demonstrate their role in drug and nutrient uptake (including in cytotoxicity or adverse drug reactions). This makes such transporters valuable both as a means of targeting drugs (not least anti-infectives) to selected cells or tissues and also as drug targets. The same considerations apply to the exploitation of substrate uptake and product efflux transporters in biotechnology. We are also beginning to recognise that transporters are more promiscuous, and antiporter activity is much more widespread, than had been realised, and that such processes are adaptive (i.e., were selected by natural evolution). The purpose of the present review is to summarise the above, and to rehearse and update readers on recent developments. These developments lead us to retain and indeed to strengthen our contention that for transmembrane pharmaceutical drug transport "phospholipid bilayer transport is negligible".

To predict human choice, consider the context

Choice prediction competitions suggest that popular models of choice, including prospect theory, have low predictive accuracy. Peterson et al. show the key problem lies in assuming each alternative is evaluated in isolation, independently of the context. This observation demonstrates how a focus on predictions can promote understanding of cognitive processes.

Integrating explanation and prediction in computational social science

Computational social science is more than just large repositories of digital data and the computational methods needed to construct and analyse them. It also represents a convergence of different fields with different ways of thinking about and doing science. The goal of this Perspective is to provide some clarity around how these approaches differ from one another and to propose how they might be productively integrated. Towards this end we make two contributions. The first is a schema for thinking about research activities along two dimensions-the extent to which work is explanatory, focusing on identifying and estimating causal effects, and the degree of consideration given to testing predictions of outcomes-and how these two priorities can complement, rather than compete with, one another. Our second contribution is to advocate that computational social scientists devote more attention to combining prediction and explanation, which we call integrative modelling, and to outline some practical suggestions for realizing this goal.