Random Deviations Improve Micro–Macro Predictions: An Empirical Test

Mäs, Michael; Helbing, Dirk

doi:10.1177/0049124117729708

Cited by 9 publications

(6 citation statements)

References 57 publications

(111 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…First, a typical characteristic of many complex systems is that even small and seemingly innocent aspects of a system can have immense impact on system behavior. In fact, theoretical as well as empirical research demonstrates that complex social systems can be in a state where even rare and random events can alter collective outcomes [83,91]. The segregation models by Schelling and Sakoda, for instance, generate higher segregation when small amounts of randomness are added to the behavior of the agents.…”

Section: Keijzer and M Mäs / The Complex Link Between Filter Bubbles ...mentioning

confidence: 99%

“…In online social networks, for instance, millions of individual users interact with a large number of network contacts, communicating content that can influence each other's opinions. Social influence between users can generate chains of reaction such that even rare idiosyncratic events can have profound impact on the system as a whole, making online social networks an ideal-typical example of complex systems [83,91,137].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

The complex link between filter bubbles and opinion polarization

Keijzer

Mäs

2022

Self Cite

View full text Add to dashboard Cite

There is public and scholarly debate about the effects of personalized recommender systems implemented in online social networks, online markets, and search engines. Some have warned that personalization algorithms reduce the diversity of information diets which confirms users’ previously held attitudes and beliefs. This, in turn, fosters the emergence opinion polarization. Critics of this personalization-polarization hypothesis argue that the effects of personalization on information diets are too weak to have meaningful effects. Here, we show that contributions to both sides of the debate fail to consider the complexity that arises when large numbers of interdependent individuals interact and exert influence on one another in algorithmically governed communication systems. Summarizing insights derived from formal models of social networks, we demonstrate that opinion dynamics can be critically influenced by mechanisms active on three levels of analysis: the individual, local, and global level. We show that theoretical and empirical research on these three levels is needed before one can determine whether personalization actually fosters polarization or not. We describe how the complexity approach can be used to anticipate and prevent undesired effects of communication technology on public debate and democratic decision-making.

show abstract

Section: Keijzer and M Mäs / The Complex Link Between Filter Bubbles ...mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

The complex link between filter bubbles and opinion polarization

Keijzer

Mäs

2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…Below, we will demonstrate that removing randomness may significantly alter the model behavior and make the mean-field approximation no longer suitable. Previous studies suggest that randomness can sufficiently change macro outcomes and, in some cases, adding randomness into the model can improve the model’s macropredictions 69 . In our case, randomness may be considered as a sort of compromise between micro and macro forecasts; whereas stochasticity suppresses the ability to predict model behavior at the microscopic level, it, apparently, gives the opportunity to achieve agreement between predictions and simulations at the macroscopic level.…”

Section: Resultsmentioning

confidence: 99%

A general framework to link theory and empirics in opinion formation models

Kozitsin

2022

Sci Rep

View full text Add to dashboard Cite

We introduce a minimal opinion formation model that is quite flexible and can reproduce a wide variety of the existing micro-influence assumptions and models. The model can be easily calibrated on real data, upon which it imposes only a few requirements. From this perspective, our model can be considered as a bridge, connecting theoretical studies on opinion formation models and empirical research on social dynamics. We investigate the model analytically by using mean-field approximation and numerically via Monte Carlo simulations. Our analysis is exemplified by recently reported empirical data drawn from an online social network. We demonstrate that the model calibrated on these data may reproduce fragmented and polarizing social systems. Furthermore, we manage to generate an artificial society that features properties quantitatively and qualitatively similar to those observed empirically at the macro scale. This ability became possible after we had advanced the model with two important communication features: selectivity and personalization algorithms.

show abstract

“…Readers are invited to conduct this test and search for explanations if they find that model outcomes change. Generally, testing effects of adding small amounts of random noise is very important in ABCM research as this may not only profoundly change model results, but sometimes also the empirical accuracy of model predictions (Macy and Tsvetkova, 2015;Mäs and Helbing, 2020). The discrete choice model also requires a new method for deciding which households get a chance to relocate.…”

Section: Table 241 Proportions Of Respondents Of Four Different Ethnic Groups In the Us (African American = Green; Asian = Blue; Hispanicmentioning

confidence: 99%

Agent-based computational models

Flache¹

2021

Research Handbook on Analytical Sociology

View full text Add to dashboard Cite

Agent-based computational modeling (ABCM) plays a central role in analytical sociology (Hedström and Bearman, 2009). Analytical sociology (AS) aims to explain social phenomena with precisely described and empirically realistic social mechanisms. Mechanistic explanations in AS target, in particular, complex micro-macro interactions in which individual behavior at the micro level of society can, in interaction with meso-and macro-level structures, bring about macro outcomes individuals neither necessarily anticipate nor desire. Examples are residential segregation (Clark and Fossett, 2008), opinion polarization (Flache, Mäs, et al., 2017), or dysfunctional status hierarchies (Mark et al., 2009). AS aspires to uncover why this occurs in real-life social settings by modeling the underlying social mechanism at a level of precision that allows logical deduction and empirical refutation of both model assumptions and predictions. With this ambition, AS is not fundamentally different from, for example, empirically oriented Rational Choice Sociology. However, despite the considerable flexibility of modern conceptualizations of Rational Choice Sociology (Wittek et al., 2013), AS is different in that is does not commit to a particular set of ingredients for modeling individual behavior (Manzo, 2014, pp. 21-27). ABCM's capacity to combine high theoretical flexibility with analytical precision is exactly what makes it so attractive to analytical sociologists.Computer simulation is the main approach in ABCM for analyzing whether, and if so, how, and under which conditions exactly the social mechanism formulated in a computational model can generate the phenomenon the researcher wants to understand (Epstein, 2006). While ABCM is not the only method for formalizing social mechanisms, it is particularly compatible with AS (Macy and Flache, 2009;Manzo, 2014Manzo, , 2021 because it combines formal precision and the ability to model and analyze complex micro-macro interactions (Mäs, Chapter 4, this volume) and theoretical flexibility in a unique way.ABCM forces a modeler to make the explanatory mechanism fully explicit. The need to translate a conceptual mechanistic model into an executable computer program requires describing algorithmically how agents arrive at changes in their actions, cognitions or emotions 1This is an open-access chapter distributed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 Unported (https://creativecommons.org/licenses/by-nc-nd/4.0/). Users can redistribute the work for non-commercial purposes, as long as it is passed along unchanged and in whole, as detailed in the License. Edward Elgar Publishing Ltd. must be clearly credited as the rights holder for publication of the original work. Any translation or adaptation of the original content requires the written authorization of Edward Elgar Publishing Ltd.The first author gratefully acknowledges financial support by the Netherlands Organization for Scientific Research (NWO) under the 2018 ORA grant ToRealSim (464.18.112). Both author...

show abstract

Random Deviations Improve Micro–Macro Predictions: An Empirical Test

Cited by 9 publications

References 57 publications

The complex link between filter bubbles and opinion polarization

The complex link between filter bubbles and opinion polarization

A general framework to link theory and empirics in opinion formation models

Agent-based computational models

Contact Info

Product

Resources

About