Social information can undermine individual performance in exploration-exploitation tasks

Yahosseini, Kyanoush Seyed; Reijula, Samuli; Molleman, Lucas; Moussaïd, Mehdi

doi:10.31234/osf.io/upv8k

Cited by 11 publications

(15 citation statements)

References 29 publications

(44 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Yet, research has mostly focused on search problems where exploration and exploitation happen simultaneously (e.g. MAB: [15]; abstract search: [16]; Lévy processes: [17]; comparison of different paradigms: [18]) and/or where jumps in the solution space are allowed (e.g., correlated MAB: [19]; sampling paradigm: [10]; secretary problem: [20]; random sampling: [21]). Nevertheless, many search problems are characterised by separated exploration and exploitation phases and gradual exploration; examples include animals deciding where to hunt prey [22,23], algorithms maximising their reward in a reinforcement learning settings [24], and humans visually searching for a lost item [25], or solving a complex problem [14,5].…”

Section: Introductionmentioning

confidence: 99%

Search as a simple take-the-best heuristic

Yahosseini

Moussaïd

2019

Preprint

Self Cite

View full text Add to dashboard Cite

Humans commonly engage in a variety of search behaviours, for example when looking for an object, a partner, information, or a solution to a complex problem. The success or failure of a search strategy crucially depends on the structure of the environment and the constraints it imposes on the individuals. Here we focus on environments in which individuals have to explore the solution space gradually and where their reward is determined by one unique solution they choose to exploit. This type of environment has been relatively overlooked in the past despite being relevant to numerous real-life situations, such as spatial search and various problem-solving tasks. By means of a dedicated experimental design, we show that the search behaviour of experimental participants can be well described by a simple heuristic model. Both in rich and poor solution spaces, a takethe-best procedure that ignores all but one cue at a time is capable of reproducing a diversity of observed behavioural patterns. Our approach, therefore, sheds lights on the possible cognitive mechanisms involved in human search.

show abstract

Section: Introductionmentioning

confidence: 99%

Search as a simple take-the-best heuristic

Yahosseini

Moussaïd

2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…activities (Lazer and Friedman, 2007;Toyokawa et al, 2014;Yahosseini et al, 2018;Özcimder et al, 2019). These studies highlight the role of the social context in exploration.…”

Section: Toward a Collective And Distributed View Of Exploration And ...mentioning

confidence: 91%

A distributed model of collective creativity in free play

Kalaydjian¹,

Laroche

Noy

et al. 2022

Front. Educ.

View full text Add to dashboard Cite

A group of children engaged in collective free play can spontaneously create new rules, learn to follow them, or find opportunities to break established ones. This rule-playing can be considered as a specific manifestation of the more general phenomenon of collective creativity. In behavioral sciences, collective creativity is often discussed as a collection of individuals each being creative. An alternative perspective views collective creativity as a distributed phenomenon: collective creativity is not a property of individual agents but rather, it emerges from the interactions within a group. Approaching free play as a case of distributed collective creativity, we understand rule-playing in terms of two complementary modes – group exploration and group exploitation-, and the transition between them. Free play is not easily amenable to fine-grained observational analysis. To overcome this, we developed the Grid Game, a new experimental setup which supports detailed empirical investigation while preserving the essence of free play. The Grid Game is a group improvisation game that uses the turn taking logic and spatial organization of typical board games, without any other predefined rules. Small groups of kids (4–5 participants) took turns in freely moving or manipulating a provided set of objects on a large 4 × 4 grid on the floor for 10 min, while being video-recorded. Despite the absence of predefined goals, simple proto-games with ad hoc rules often emerge, for example, placing objects on top of each other (create a tower) or an aiming-to-a-target shooting game. We propose an analysis of the emergence of such proto-games in terms of group exploration and exploitation. Building on our previous work on the Creative Foraging Game, we focus on cases of transition from exploration to exploitation underlying the discovery of a new form or rule. Based on Choreographer João Fiadeiro’s body of work, we describe these phase shifts as a distributed process composed of three stages: (1) Suggestion, (2) Recognition, and (3) Confirmation. We provide detailed descriptions of game moments according to this model which demonstrate the distributed nature of collective creativity in free play.

show abstract

“…Collective problem solving involves a fundamental tension: the same factor -communication -pulls in opposite directions. The benefit is that social learning spreads information necessary to solve a particular problem; the cost is that this simultaneously reduces both the informational diversity in the group (Lazer & Friedman, 2007) and the group's tendency to find new information (Yahosseini, Reijula, Molleman, & Moussaïd, 2018). Social learning can also cause a group to converge on something because it is popular, rather than because it is good, making outcomes increasingly unpredictable (Salganik, Dodds, & Watts, 2006).…”

Section: Diversity In Social Learning Strategies Resolves An Underlying Tensionmentioning

confidence: 99%

The Diversity Gap: When Diversity Matters for Knowledge

Sulik

Bahrami

Deroy

2021

Perspect Psychol Sci

View full text Add to dashboard Cite

Can diversity make for better science? Although diversity has ethical and political value, arguments for its epistemic value require a bridge between normative and mechanistic considerations, demonstrating why and how diversity beneﬁts collective intelligence. However, a major hurdle is that the beneﬁts themselves are rather mixed: Quantitative evidence from psychology and behavioral sciences sometimes shows a positive epistemic effect of diversity, but often shows a null effect, or even a negative effect. Here we argue that to make progress with these why and how questions, we need ﬁrst to rethink when one ought to expect a beneﬁt of cognitive diversity. In doing so, we highlight that the beneﬁts of cognitive diversity are not equally distributed about collective intelligence tasks and are best seen for complex, multistage, creative problem solving, during problem posing and hypothesis generation. Throughout, we additionally outline a series of mechanisms relating diversity and problem complexity, and show how this perspective can inform metascience questions.

show abstract

Social information can undermine individual performance in exploration-exploitation tasks

Cited by 11 publications

References 29 publications

Search as a simple take-the-best heuristic

Search as a simple take-the-best heuristic

A distributed model of collective creativity in free play

The Diversity Gap: When Diversity Matters for Knowledge

Contact Info

Product

Resources

About