Reference-dependent preferences arise from structure learning

Le, Hunter; Gershman, Samuel J.

doi:10.1101/252692

Cited by 12 publications

(12 citation statements)

References 54 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Most empirical studies of between-choice non-multiattribute reference effects have focused on decision from description (Rigoli et al, 2016a; 2016b; 2016c; 2018; Stewart et al, 2015;Walasek & Stewart, 2015). However, we note also studies about decision from experience which have manipulated the contextual average(Hunter & Gershman, 2018;Rigoli et al, 2018). Results of these studies are analogous to those emerged during decision from description.…”

supporting

confidence: 55%

“…EaR models rely on a perspective analogous to PT (Hunter & Gershman, 2018;Kőszegi & Rabin, 2006; Rigoli et al, 2016a; 2017). However, they do not interpret the frame of reference as the status quo, but as reflecting expectations about upcoming stimuli (Kőszegi & Rabin, 2006).…”

Section: Theoriesmentioning

confidence: 99%

“…However, an explicit and systematic treatment of reference effects was first proposed by a psychological account, namely Prospect Theory (PT; Camerer et al, 2011; Kahneman & Tversky, 1979), where the concept of reference point (intended as the status quo) plays a prominent role. Building on this and analogous ideas, theoretical and empirical work has probed reference effects further, for example examining the role of expectations (Hunter & Gershman, 2018;Kőszegi & Rabin, 2006;Rigoli et al, 2016a; 2017) memory retrieval (Brown & Matthews, 2011;Stewart, 2009;Stewart et al, 2006;2015), and efficient coding Simonson & Tversky, 1992;Soltani et al, 2012;Trueblood et al, 2014;Tsetsos et al, 2010;Tversky, 1972).…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Reference effects on decision-making elicited by previous rewards

Rigoli

2019

Cognition

View full text Add to dashboard Cite

Substantial evidence has highlighted reference effects occurring during decision-making, whereby subjective value is not calculated in absolute terms but relative to the distribution of rewards characterizing a context. Among these, within-choice effects are exerted by options simultaneously available during choice. These should be distinguished from between-choice effects, which depend on the distribution of options presented in the past. Influential theories on between-choice effects include Decision-by-Sampling, Expectation-as-Reference and Divisive Normalization. Surprisingly, previous literature has focused on each theory individually disregarding the others. Thus, similarities and differences among theories remain to be systematically examined. Here we fill this gap by offering an overview of the state-of-the-art of research about between-choice reference effects. Our comparison of alternative theories shows that, at present, none of them is able to account for the full range of empirical data. To address this, we propose a model inspired by previous perspectives and based on a logistic framework, hence called logistic model of subjective value. Predictions of the model are analysed in detail about reference effects and risky decision-making. We conclude that our proposal offers a compelling framework for interpreting the multifaceted manifestations of between-choice reference effects.

show abstract

supporting

confidence: 55%

Section: Theoriesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reference effects on decision-making elicited by previous rewards

Rigoli

2019

Cognition

View full text Add to dashboard Cite

show abstract

“…In summary, our results highlight the highly dynamic and rational nature of value representation and resolve the tension between optimal-but-costly absolute value encoding, and efficient-but-potentially-irrational, relative value encoding. Humans do not simply have a single, fixed form of representation, but rather adjust their code in a resource-efficient [45][46][47] manner according to expected task demands. Further, our findings highlight that both absolute and relative codes previously found can be explained by the fact that participants infer which code would be sufficient for the current task.…”

Section: Discussionmentioning

confidence: 99%

Human value learning and representation reflects rational adaptation to task demands

Juechems¹,

Altun²,

Hira³

et al. 2021

Preprint

View full text Add to dashboard Cite

When making decisions about goods and actions, humans and animals often rely on internally represented values. However, to be useful across a wide range of contexts, these values need to be represented on an absolute scale – a coding scheme that is computationally costly. By contrast, representing values in a way that depends entirely on context is highly computationally efficient, but can lead to irrational behaviour when values need to be compared across contexts. Thus, an efficient learner would allocate limited computational resources only when needed according to their expectations about the future. Here, we test the hypothesis that value representation is not fixed, but rationally adapted to expectations in two human learning experiments. Unlike most lab-based tasks, participants could use their initial experience (Phase 1) to optimise behaviour (Phase 2). Phase 1 was designed to cause one group to expect only decisions within local contexts (relative code sufficient), and another group to expect choices across local contexts (relative code insufficient). Despite identical learning experiences, the group whose expectations included choices across local contexts, went on to learn absolute value representations, and learned more absolute-like representations than the other group. Human value representation is neither relative nor absolute, but efficiently and rationally tuned to task demands.

show abstract

“…The idea that preferences adapt to fit a given distribution is neither new nor unfounded (Brunswik, 1956;Gigerenzer et al, 1991;Glöckner et al, 2014;Weber & Johnson, 2008). Indeed, while prospect theory rests on reference-dependence, several newer models mimic RDU in that they claim that the values with which we imbue our options rely on the other options we have at our disposal (Hunter & Gershman, 2018;Loomes & Sugden, 2006;Parducci, 2012;Steward et al, 2003;Yaari, 2006). Likewise, it has long been known in psychology and neuroscience that distribution-adaptation is an inherent feature of the brain (Louie & De Martino, 2013).…”

Section: Discussionmentioning

confidence: 99%

Adaptation of utility functions to reward distribution in rhesus monkeys

Bujold

Ferrari-Toniolo

Schultz

2020

Preprint

View full text Add to dashboard Cite

30This study investigated the influence of experienced reward distributions on the shape of utility 31 functions inferred from economic choice. Utility is the hypothetical variable that appears to be max-32 imized by the choice. Despite the generally accepted notion that utility functions are not insensitive 33 to external references, the exact occurrence of such changes remains largely unknown. Here we 34 benefitted from the capacity to perform thorough and extensive experimental tests of one of our 35 evolutionary closest, experimentally viable and intuitively understandable species, the rhesus ma-36 caque monkey. Data from thousands of binary choices demonstrated that the animals' preferences 37 changed dependent on the statistics of recently experienced rewards and adapted to future expected 38 rewards. The elicited utility functions shifted and extended their shape with several months of 39 changes in the mean and range of reward distributions. However, the adaptations were usually not 40complete, suggesting that past experiences remained present when anticipating future rewards. 41Through modelling, we found that reinforcement learning provided a strong basis for explaining 42 these adaptations. Thus, rather than having stable and fixed preferences assumed by normative eco-43 nomic models, rhesus macaques flexibly shaped their preferences to optimize decision-making ac-44 cording to the statistics of the environment. 45 46 109 far from being fixed and abstract, preferences follow the expectation of what animals think might 110 happen given the knowledge they have accumulated over time.

show abstract

Reference-dependent preferences arise from structure learning

Cited by 12 publications

References 54 publications

Reference effects on decision-making elicited by previous rewards

Reference effects on decision-making elicited by previous rewards

Human value learning and representation reflects rational adaptation to task demands

Adaptation of utility functions to reward distribution in rhesus monkeys

Contact Info

Product

Resources

About