A decision network account of reasoning about other people’s choices☆
Introduction
People tend to assume that other people’s behavior results from their conscious choices—for example, choices about what outfit to wear, what movie to watch, or how to respond to a question (Gilbert and Malone, 1995, Ross, 1977). Reasoning about choices like these requires an understanding of how they are motivated by mental states, such as what others know and want. Even though mental states are abstract entities and are inaccessible to other people, most people find it natural to make predictions about what others will choose and infer why they made the choices they did. In this paper, we explore the computational principles that support such inferences.
Several models of how people reason about others’ actions have been proposed by social and developmental psychologists (e.g., Gilbert, 1998, Jones and Davis, 1965, Malle and Knobe, 1997, Wellman and Bartsch, 1988). Four examples are shown in Fig. 1. For instance, Fig. 1a shows a model of how people reason about others’ actions. According to this model, a person’s dispositional characteristics combine to produce an intention to take an action. Then, if that person has the necessary knowledge and ability to carry out the action, he or she takes the action, producing a set of effects. The other three models in Fig. 1 take into account additional variables such as belief and motivation. For example, Fig. 1b proposes that people take actions that they believe will satisfy their desires. The models in Fig. 1 are not computational models, but they highlight the importance of structured causal representations for reasoning about choices and actions. We show how representations like these can serve as the foundation for a computational account of social reasoning.
Our account draws on two general themes from the psychological literature. First, people tend to assume that choices, unlike world events, are goal-directed (Baker et al., 2009, Csibra and Gergely, 1998, Csibra and Gergely, 2007, Goodman et al., 2009, Shafto et al., 2012). We refer to this assumption as the principle of goal-directed choice, although it has also been called the intentional stance (Dennett, 1987) and the principle of rational action (Csibra and Gergely, 1998, Csibra and Gergely, 2007). Second, we propose that human reasoning relies on probabilistic inference. Recent work on inductive reasoning has emphasized the idea that probabilistic inference can be carried out over structured causal representations (Griffiths and Tenenbaum, 2005, Griffiths et al., 2010, Tenenbaum et al., 2011), and our account relies on probabilistic inference over representations similar to those in Fig. 1.
Our account is related to previous work on Bayes nets (short for Bayesian networks; Pearl, 2000), which have been widely used to account for causal reasoning (Gopnik et al., 2004, Sloman, 2005). Bayes nets rely on probabilistic inference over structured causal representations, but they do not capture the principle of goal-directed choice. In this paper, we present an extension of Bayes nets called decision networks1 (Howard and Matheson, 2005, Russell and Norvig, 2010) that naturally captures the principle of goal-directed choice. We propose that people reason about choice behavior by constructing mental models of other people’s choices that are similar to decision networks and then performing probabilistic inference over these mental models. Decision networks may therefore provide some computational substance to qualitative models like the ones in Fig. 1.
Our decision network account of reasoning about choices builds on previous approaches, including the theory theory of conceptual structure. The theory theory proposes that children learn and reason about the world by constructing scientific-like theories that are testable and subject to revision on the basis of evidence (Gopnik & Wellman, 1992). These theories can exist at different levels of abstraction. Framework theories capture fundamental principles that are expected to apply across an entire domain, and these framework theories provide a basis for constructing specific theories of concrete situations (see Wellman & Gelman, 1992). The decision network account can be viewed as a framework theory that captures the idea that choices are made in order to achieve goals, whereas individual decision networks can be viewed as specific theories. Gopnik and Wellman (2012) argue that Bayes nets provide a way to formalize the central ideas of the theory theory, and their reasons apply equally well to decision networks. For example, decision networks can be used to construct abstract causal representations of the world, to predict what will happen next, or to infer unobserved causes.
Although decision networks have not been previously explored as psychological models, they have been used by artificial intelligence researchers to create intelligent agents in multi-player games (Gal and Pfeffer, 2008, Koller and Milch, 2003, Suryadi and Gmytrasiewicz, 1999). In the psychological literature there are a number of computational accounts of reasoning about behavior (Bello and Cassimatis, 2006, Bonnefon, 2009, Bonnefon and Sloman, 2013, Hedden and Zhang, 2002, Oztop et al., 2005, Shultz, 1988, Van Overwalle, 2010, Wahl and Spada, 2000), and some of these accounts rely on Bayes nets (Goodman et al., 2006, Hagmayer and Osman, 2012, Hagmayer and Sloman, 2009, Sloman and Hagmayer, 2006, Sloman et al., 2012). However, our work is most closely related to accounts that extend the framework of Bayes nets to include the principle of goal-directed choice (Baker and Tenenbaum, 2014, Baker et al., 2008, Baker et al., 2009, Baker et al., 2011, Doshi et al., 2010, Goodman et al., 2009, Jara-Ettinger et al., 2012, Pantelis et al., 2014, Pynadath and Marsella, 2005, Tauber and Steyvers, 2011, Ullman et al., 2009). Much of this work uses a computational framework called Markov decision processes (MDPs; Baker and Tenenbaum, 2014, Baker et al., 2009).
In the next section, we describe the decision network framework in detail and explain how it is related to the Bayes net and MDP frameworks. We then present four experiments that test predictions of the decision network framework as an account of how people reason about choice behavior. Our first two experiments are specifically designed to highlight unique predictions of decision networks that distinguish them from an account based on standard Bayes nets. Our second two experiments focus on inferences about mental states. Experiment 3 focuses on inferences about what someone else knows and Experiment 4 focuses on inferences about what someone else’s goals are.
Section snippets
Decision networks
We will introduce the details of decision networks (decision nets for short) with the following running example. Suppose Alice is playing a game. In the game, a two-colored die is rolled, and if Alice chooses the color of the rolled die, she earns a reward. Suppose further that Alice is able to see the outcome of the rolled die before making her choice. This situation can be represented using the decision net in Fig. 2.
Decision nets distinguish between four different kinds of variables: world
Overview of experiments
We conducted four experiments to evaluate how well the decision net framework accounts for people’s inferences about choice behavior. Our first two experiments were designed to directly compare decision nets with Bayes nets that do not contain a notion of goal-directed choice. Experiment 1 focuses on predicting choices after a utility function changes, and Experiment 2 focuses on using observed choices to make causal inferences. Our second two experiments examine how people make inferences
Experiment 1: Predicting other people’s choices
Given that it is possible to compile any decision net into a Bayes net that makes identical choice predictions, it is important to ask whether the decision net is a better psychological model than the compiled Bayes net version of the same network. Earlier, we argued that one advantage of a decision net over a Bayes net is that a decision net can naturally accommodate changes to the utility function. When the utility function of a decision net changes, the decision net predicts that the
Experiment 2: Reasoning about goal-directed choices
The purpose of Experiment 2 was to explore whether people rely on the assumption of goal-directed choice even when they are not asked to make direct judgments about others’ choices. We asked participants to make a causal inference that was informed by someone’s choice. Specifically, we asked participants to observe another person play a single round of the cruise ship game and then infer which machine the player was using. Such inferences are possible by considering how much money a player
Experiment 3: Inferring what other people know
We have hypothesized that people reason about other people’s choices by constructing mental models of those choices and that these mental models are similar to decision nets. Our second two experiments further test this hypothesis by focusing on two more inferences from Table 1 that are related to choice behavior. In particular, Experiments 3 and 4 focus on inferences about mental states. Mental state inferences are common in social interactions. For example, when you see a man gossiping about
Experiment 4: Inferring other people’s goals
Experiment 3 showed that performing model selection over decision nets accounted well for people’s inferences about what someone else knows. The purpose of Experiment 4 was to apply this same account to a situation in which people must infer someone else’s goals.
General discussion
We proposed that decision networks can help to understand how people reason about the choices of others. Decision nets are extensions of Bayes nets, and the critical difference between the two is that decision nets incorporate the assumption of goal-directed choice. The results of our first two experiments suggested that people make use of this assumption when reasoning about other people’s choices. Experiment 1 showed that this assumption allows people to quickly adjust their predictions about
Conclusion
We presented a computational framework for reasoning about choices. We proposed that people reason about other people’s choices by constructing mental models that are similar to decision nets. Decision nets use structured causal representations and incorporate two key computational principles: goal-directed choice and probabilistic inference. Decision net models can be used to make predictions and inferences about other people’s choices, and we described four experiments in which decision nets
References (99)
- et al.
Action understanding as inverse planning
Cognition
(2009) - et al.
Stochastic dynamic programming with factored representations
Artificial Intelligence
(2000) Mental models and counterfactual thoughts about what might have been
Trends in Cognitive Sciences
(2002)- et al.
‘Obsessed with goals’: Functions and mechanisms of teleological interpretation of actions in humans
Acta Psychologica
(2007) - et al.
Statistically optimal perception and learning: From behavior to neural representations
Trends in Cognitive Sciences
(2010) - et al.
Probabilistic models of cognition: Exploring representations and inductive biases
Trends in Cognitive Sciences
(2010) - et al.
What do you think I think you think?: Strategic reasoning in matrix games
Cognition
(2002) - et al.
From influence diagrams to junction trees
- et al.
A probabilistic account of exemplar and category generation
Cognitive Psychology
(2013) Mental models in cognitive science
Cognitive Science
(1980)
Multi-agent influence diagrams for representing and solving games
Games and Economic Behavior
Developing a concept of choice
The folk concept of intentionality
Journal of Experimental Social Psychology
Dispositional inference from effects of actions: Effects chosen and effects forgone
Journal of Experimental Social Psychology
Mental state inference using visual control parameters
Cognitive Brain Research
Inferring the intentional states of autonomous virtual agents
Cognition
The intuitive psychologist and his shortcomings: Distortions in the attribution process
Causal structure learning over time: Observations and interventions
Cognitive Psychology
The causal psycho-logic of choice
Trends in Cognitive Sciences
Inferring causal networks from observations and interventions
Cognitive Science
Inferential processes in the forced compliance situation: A Bayesian analysis
Journal of Experimental Social Psychology
Infants’ teleological and belief inference: A recurrent connectionist approach to their minimal representational and computational requirements
NeuroImage
Young children’s reasoning about beliefs
Cognition
A Bayesian analysis of attribution processes
Psychological Bulletin
Uniqueness of behavioral effects in causal attribution
Journal of Personality
Modeling human plan recognition using Bayesian theory of mind
The curse of knowledge in reasoning about false belief
Psychological Science
A theory of utility conditions: Paralogical reasoning from decision-theoretic leakage
Psychological Review
The causal structure of utility conditionals
Cognitive Science
Decision-theoretic planning: Structural assumptions and computational leverage
Journal of Artificial Intelligence Research
The teleological origins of mentalistic action explanations: A developmental hypothesis
Developmental Science
The intentional stance
Networks of influence diagrams: A formalism for representing agents’ beliefs and decision-making processes
Journal of Artificial Intelligence Research
Mental models as representations of discourse and text
Ordinary personology
The correspondence bias
Psychological Bulletin
In defense of simulation theory
Mind & Language
A theory of causal learning in children: Causal maps and Bayes nets
Psychological Review
Bayesian networks, Bayesian learning and cognitive development
Developmental Science
Why the child’s theory of mind really is a theory
Mind & Language
Reconstructing constructivism: Causal models, Bayesian learning mechanisms and the theory theory
Psychological Bulletin
Cited by (21)
Naïve information aggregation in human social learning
2024, CognitionInteractive cognitive maps support flexible behavior under threat
2023, Cell ReportsThe computational challenge of social learning
2021, Trends in Cognitive SciencesCitation Excerpt :That these factors also interact with one another, further compounds the challenge of modeling how learning unfolds in the social world. The first challenge for modeling the social learning problem is the fact that another person’s mood, motives, and intentions, which we collectively refer to as their internal state (see Glossary), are not observable and are therefore highly uncertain [27,28]. This poses a problem for computing the values of our own potential actions, since they depend on our partners’ internal state (Figure 2, purple box) and how our partner will respond to our actions in turn (Figure 2, orange box).
A computational framework for understanding the roles of simplicity and rational support in people's behavior explanations
2021, CognitionCitation Excerpt :First, behavior explanations are more likely to refer to mental states like beliefs and desires that gave rise to an intention to act (Malle, 1999). Second, when reasoning about other people's mental states, people expect others to behave in a goal-directed way (Baker, Jara-Ettinger, Saxe, & Tenenbaum, 2017; Baker, Saxe, & Tenenbaum, 2009; Baker & Tenenbaum, 2014; Jara-Ettinger, Gweon, Schulz, & Tenenbaum, 2016; Jern & Kemp, 2015; Ullman et al., 2010), an expectation that can even influence their causal judgments (Kirfel & Lagnado, 2019; Lagnado & Channon, 2008). This observation leads to the rational support principle.
The relational logic of moral inference
2021, Advances in Experimental Social PsychologyCitation Excerpt :Our recent work suggests an additional computational goal for moral inference: learners should seek to form accurate beliefs, but also should update those beliefs in ways that facilitate the development and maintenance of social relationships. Recent studies in cognitive science have sought to investigate the cognitive mechanisms through which people infer other's hidden preferences, intentions, and desires over time (Aksoy & Weesie, 2014; Diaconescu et al., 2014; Jern & Kemp, 2015). By measuring the behaviors of individual learners and then fitting descriptive models to learners' behavior, researchers can determine the extent to which behavior conforms to the predictions of the Bayesian ideal.
People learn other people's preferences through inverse decision-making
2017, CognitionCitation Excerpt :Naïve utility calculus refers to the expectation people have that others will generally make choices that produce greater utility. Combining naïve utility calculus with inverse reasoning has led to a number of useful accounts of social inference in recent years (Baker, Jara-Ettinger, Saxe, & Tenenbaum, 2017; Baker, Saxe, & Tenenbaum, 2009; Baker & Tenenbaum, 2014; Jern & Kemp, 2015; Tauber & Steyvers, 2011; Ullman et al., 2009; Wu, Baker, Tenenbaum, & Schulz, 2014). However, few studies in this literature have explored the basic question of how people infer what other people like and dislike by observing their choices.
- ☆
Data from Experiments 3 and 4 were presented at the 33rd Annual Conference of the Cognitive Science Society. We thank Jessica Lee for helping to collect the data for Experiments 1 and 2. We thank David Danks for feedback on the development of this work, and Jean-François Bonnefon, Mark Steyvers, and two anonymous reviewers for feedback on the manuscript. This work was supported by the Pittsburgh Life Sciences Greenhouse Opportunity Fund and by the National Science Foundation (NSF) Grant CDI-0835797. Alan Jern was supported in part by NIMH Training Grant T32MH019983.