© Copyright JASSS

JASSS logo ------

Computational Social Psychology

Robin R. Vallacher
Routledge: London, 2017
ISBN

Order this book

Reviewed by Wander Jager
University of Groningen

Cover of book Introduction

The field of social psychology originally - almost a century ago - was very much interested in the complex interactions between people. Due to lack of experimental control, much research within this field has taken the controlled laboratory experiment as the dominant tool to study social psychological phenomena. In their edited book, Vallacher, Read and Nowak also signal the problems associated with this methodology, in particular the issue that “verbal” theories based on statistical analysis do not address the causal mechanisms, and the problems of addressing behavioural dynamics in groups. Computational social psychology is being presented in the book as the methodological approach that is capable of bringing social psychology to the next level of scientific discovery.

Social psychology is an important field in understanding the behavioural dimensions of many societal challenges and is sometimes considered as a “hub science” connecting different other disciplines. As such the development of computational models is not only important within the field of social psychology, but also highly relevant to connect psychological theories to other scientific disciplines in order to address multidisciplinary challenges, such as adaptation to climate change, opinion dynamics, transitions in energy use and food consumption, and the adaptive capacity of societies dealing with migration (see, e.g., Jager 2017).

This edited book is aimed at providing an in-depth view of recent developments within computational social psychology. Following an introductory chapter, the book is organised in 3 parts, respectively addressing dynamics at the intrapersonal, interpersonal and collective level. The book concludes with three chapters discussing perspectives on transforming social psychology. In this review, I will first guide you through the respective chapters. Following that, I will reflect on the potential readership of this book, and how to use the book.

Introductory chapter

The first introductory chapter clearly describes the key argument of how computation opens a new and much-needed paradigm in social psychology by offering a tool to test formal rules describing social psychological processes. Whereas much of the history in social psychological experimentation has been based on a statistical approach, computational social psychology offers a methodology to study the possible causal mechanisms of phenomena. Obviously, the main advantage of computational models is the possibility to study the dynamics of social psychological processes, which offers a new powerful paradigm to study human behaviour. This is an important development, considering, in particular, the problems the old social psychology paradigm is facing concerning lack of replication, artificial and non-representative laboratory settings, and the fundamental limitations of statistical analysis (Open Science Confederation, 2015). Computational social psychology is considered by several scholars to be a key methodology to bring the field of psychology to the next level of scientific rigour.

Part 1 of the book: the intra-personal level

Chapter 2 on artificial personalities gives a clear demonstration of how computational methods can bring together different perspectives on personality, and demonstrate how the big 5 of personality can merge, and how personality can operate as a dynamic construct depending on external conditions.

Chapter 3 addresses impression formation using connectionist models. Thought experiments are discussed in deep detail on for example order effects in impression formation. Then simulations are discussed demonstrating the principle. For a social simulator it would have been good to get some more information on the model, or even (pseudo)code.

Chapter 4 explains how psychology became a very fractioned discipline. Moreover, it is stated that a holist perspective on humans is needed to contribute with a psychological theory to the complexity perspective. An example is given of an attachment neural network model that models different processes that play a critical role in social interaction. A plea is made for integrating cognition, emotion, and motivation, and considers them in developmental and social interaction contexts.

Chapter 5 addresses computational models of health behaviour and starts with the explanation of a system dynamics approach to the Theory of Planned Behaviour. Next, an artificial neural network is being used to model the Theory of Reasoned Action in a health behaviour context. Next, a multi-agent neural network is being used to explore dynamics and heterogeneity in populations, demonstrating that the diffusion of intentions is very susceptible for the assumptions made in the model, and thus a challenge for computational social science.

Part 2 of the book: the interpersonal level

In Chapter 6 a perspective is given on the different time dimensionalities in human behaviour (from the nervous system to culture), and how social synergies are established, maintained and organised across time. A very interesting and relevant fractal scaling of human social behaviour is being discussed, where social behaviour can be considered as a contextually constrained property of a complex social environment instead of a property of individuals.

Chapter 7 addresses the emotional interactions between people using an equation based model. A case is presented where romantic partners are interacting with their respective body weights (BMI). The approach demonstrates that computational methods are suitable to study emotional interactions between people in a novel way. In the discussion, the authors reflect on the possibilities for using this method in larger groups of interacting people and on other topics.

Chapter 8 addresses co-regulation in social relations, using a Dynamical Systems Model to study the relation between the heart rates of partners. This chapter explains the method in deep detail; however, the empirical setting responsible for synchronicities in heart rate remained to be guessed.

Chapter 9 discusses a self-organisational perspective on coordination between people, also elaborating on the principle of synchronicity. This is based on an evolutionary perspective on how organisms collaborate and operate in an environment. A key contribution for social psychology is the development of a perspective on understanding collaboration that goes beyond the study of individual behaviour/motives (or dyads).

Also, chapter 10 focuses on synchronisation of behaviour, and explains logistic equations as an approach to simulate human synchronisation in a simple way. The authors emphasize the importance of using as simple as possible models to simulate behaviour, a stand that may be debated.

Part 3 of the book: collective dynamics in social systems

Chapter 11 is exploring how different network roles may affect the goal striving in individuals. Focussing on the example of overeating/obesity, the authors describe how different social interactions over time with different people may result in a dynamic change of the goal to reduce weight.

Chapter 12 continues with the discussion of computational models of social influence and collective behaviour. This chapter focuses on the burden of social proof (BOP) to use in quantitative modelling. It is demonstrated that such an approach is useful to quantitatively describe classical experiments such as of Asch and Milgram.

Chapter 13 is discussing the issue of modelling cultural dynamics and is addressing this from the approach of multidimensional public opinion dynamics and the rise and fall of cooperation in groups (evolutionary game theory). Both approaches have a different background but share the modelling perspective. In the discussion, the questions addressed if their complementarity can result in a synthesis between the approaches.

The concluding part of the book: the transformation of social psychology

Chapter 14 opens with a nice example of why formal modelling is important to improve our often limited understanding of social phenomena, and why simple models are very useful. Also it is being explained that many exiting social psychological theories are of a verbal nature, which makes falsification a problem. This is an important reason to work towards the formalisation of social psychological theories. Some examples are given – starting with Newton’s gravitational model – to explain why simple models can contribute to scientific progress.

Chapter 15 addresses big data in social scientific research. It is being explained that big data is changing social scientific research concerning the type of research design being used, the new opportunities and challenges that big data offer for the social sciences. A number of examples are given of data derived from platforms, and one detailed study on Twitter data is being presented. However, these are not being linked to computational models, and the theoretical contribution is not being elaborated upon, which is also identified as a challenge. The main question that remains is to what extent big data is capable of representing “real” human behaviour instead of data on what people report on, e.g., Twitter and Facebook.

Chapter 16, the concluding chapter, is reflecting on the future of computational social psychology. Partly the discussion addresses the large number of data that will become available, e.g., through the internet of things, and the capacity to analyse this. Also, attention is given to agent based simulation and the possibility to work towards integrated social psychological models.

Conclusions

Overall this book clearly explains the relevance of computational methods for the development of a more causal and dynamical social psychology and illustrates that with a variety of applications. Being a social psychologist myself working with computational models for many years I was already convinced of the relevance of computational models for the field. I assume much of the readership of JASSS is also aware of the importance of computational methods to study behavioural dynamics. As such I assume that for the JASSS readership the different chapters addressing various implementations of computational models are of specific interest, as they may provide examples of how to implement different theoretical insights. In my view, most of the chapters provide a more generic description of how to develop computational models, and the reader is not presented with, e.g., pseudo-code or direct references to running models that can be implemented easily. Rather, the reader interested in a particular phenomenon/theory can use relevant chapters to get an overview of the approach, and a good listing of references for getting further acquainted with this particular computational approach. Hence I guess this book is a valuable referential book to have in every library being used by computational social scientists.

In my view, this book is not meant as a generic introduction for non-computational social psychologists. Convincing students in social psychology would, for example, require a broader introduction into computational social science, principles of behavioural dynamics and complexity, data and validation and a complete representation of scientific work using (social)psychological theory in computational settings. The readership of JASSS will recognise that there are many more social computational approaches being published and used that rely on (social) psychological theory. Challenges in modelling these social processes are recognised and explicitly being addressed, e.g., by Flache et al. (2007). Also, several research groups work on the development of more integrated models – also advocated in this book as an important development – formalising the connections between human (social) cognition, motivation, decision-making, interaction and behaviour in different contexts.

I hope this book will serve as a pillar in further bridging the divide between computational scientists and social psychologists. Considering the challenges we face in modelling human behaviour in a valid way to deepen our understanding of many societal challenges, it is relevant to develop sound social psychological computational models. As such, I recommend this book as an important reference book for modellers to keep updated on the latest developments in the computational modelling of human behaviour.


* References

FLACHE, A., Mäs, M., Feliciani, T., Chattoe-Brown, E., Deffuant, G., Huet, S. and Lorenz, J., (2017). Models of Social Influence: Towards the Next Frontiers'. Journal of Artificial Societies and Social Simulation 20 (4) 2 http://jasss.soc.surrey.ac.uk/20/4/2.html. doi: 10.18564/jasss.352

JAGER, W. (2017). Enhancing the Realism of Simulation (EROS): On Implementing and Developing Psychological Theory in Social Simulation. Journal of Artificial Societies and Social Simulation 20 (3) 14 http://jasss.soc.surrey.ac.uk/20/3/14.html. doi: 10.18564/jasss.3522

OPEN SCIENCE COLLABORATION. (2015). Estimating the reproducibility of psychological science. Science, 349(6251), aac4716. Doi: 10.1126/science.aac4716

-------

ButtonReturn to Contents of this issue