Riccardo Boero, Economics and Market Analysis Team, Energy and Infrastructure Analysis Group, Los Alasmos National Laboratory, USA

Select Guide Rating

• Preface ix1 Introduction: Toward behavioral computational social science 11.1 Research strategies in CSS 21.2 Why behavioral CSS 31.3 Organization of the book 4Part i CONCEPTS AND METHODS 72 Explanation in computational social science 92.1 Concepts 102.1.1 Causality 102.1.2 Data 182.2 Methods 192.2.1 ABMs 192.2.2 Statistical mechanics, system dynamics, and cellular automata 222.3 Tools 252.4 Critical issues: Uncertainty, model communication 273 Observation and explanation in behavioral sciences 313.1 Concepts 323.2 Observation methods 353.2.1 Naturalistic observation and case studies 353.2.2 Surveys 363.2.3 Experiments and quasiexperiments 373.3 Tools 383.4 Critical issues: Induced responses, external validity, and replicability 404 Reasons for integration 434.1 The perspective of agent]based modelers 444.2 The perspective of behavioral social scientists 494.3 The perspective of social sciences in general 54Part iI BEHAVIORAL COMPUTATIONAL SOCIAL SCIENCE IN PRACTICE 575 Behavioral agents 595.1 Measurement scales of data 615.2 Model calibration 635.2.1 Single decision variable and simple decision function 635.2.2 Multiple decision variables and multilevel decision trees 655.3 Model classification 675.4 Critical issues: Validation, uncertainty modeling 706 Sophisticated agents 736.1 Common features of sophisticated agents 756.2 Cognitive processes 756.2.1 Reinforcement learning 766.2.2 Other models of bounded rationality 806.2.3 Nature]inspired algorithms 806.3 Cognitive structures 846.3.1 Middle]level structures 856.3.2 Rich cognitive models 866.4 Critical issues: Calibration, validation, robustness, social interface 887 Social networks and other interaction structures 917.1 Essential elements of SNA 937.2 Models for the generation of social networks 997.3 Other kinds of interaction structures 1047.4 Critical issues: Time and behavior 1068 An example of application 1098.1 The social dilemma 1108.1.1 The theory 1118.1.2 Evidence 1138.1.3 Our research agenda 1148.2 The original experiment 1148.3 Behavioral agents 1168.3.1 Fixed effects model 1168.3.2 Random coefficients model 1178.3.3 First differences model 1188.3.4 Ordered probit model with individual dummies 1198.3.5 Multilevel decision trees 1218.3.6 Classified heuristics 1268.4 Learning agents 1278.5 Interaction structures 1278.6 Results: Answers to a few research questions 1288.6.1 Are all models of agents capable of replicating the experiment? 1298.6.2 Was the experiment influenced by chance? 1318.6.3 Do economic incentives work? 1338.6.4 Why does increasing group size generate more cooperation? 1358.6.5 What happens with longer interaction? 1368.6.6 Does a realistic social network promote cooperation? 1378.7 Conclusions 138Appendix Technical guide to the example model 141A.1 The interface 142A.2 The code 145A.2.1 Variable declaration 146A.2.2 Simulation setup 152A.2.3 Running the simulation 157A.2.4 Decision-making 157A.2.5 Updating interaction structure and other variables 165References 173Index 187

Riccardo Boero, Economics and Market Analysis Team, Energy and Infrastructure Analysis Group, Los Alasmos National Laboratory, USA

Select Guide Rating

• Preface ix1 Introduction: Toward behavioral computational social science 11.1 Research strategies in CSS 21.2 Why behavioral CSS 31.3 Organization of the book 4Part i CONCEPTS AND METHODS 72 Explanation in computational social science 92.1 Concepts 102.1.1 Causality 102.1.2 Data 182.2 Methods 192.2.1 ABMs 192.2.2 Statistical mechanics, system dynamics, and cellular automata 222.3 Tools 252.4 Critical issues: Uncertainty, model communication 273 Observation and explanation in behavioral sciences 313.1 Concepts 323.2 Observation methods 353.2.1 Naturalistic observation and case studies 353.2.2 Surveys 363.2.3 Experiments and quasiexperiments 373.3 Tools 383.4 Critical issues: Induced responses, external validity, and replicability 404 Reasons for integration 434.1 The perspective of agent]based modelers 444.2 The perspective of behavioral social scientists 494.3 The perspective of social sciences in general 54Part iI BEHAVIORAL COMPUTATIONAL SOCIAL SCIENCE IN PRACTICE 575 Behavioral agents 595.1 Measurement scales of data 615.2 Model calibration 635.2.1 Single decision variable and simple decision function 635.2.2 Multiple decision variables and multilevel decision trees 655.3 Model classification 675.4 Critical issues: Validation, uncertainty modeling 706 Sophisticated agents 736.1 Common features of sophisticated agents 756.2 Cognitive processes 756.2.1 Reinforcement learning 766.2.2 Other models of bounded rationality 806.2.3 Nature]inspired algorithms 806.3 Cognitive structures 846.3.1 Middle]level structures 856.3.2 Rich cognitive models 866.4 Critical issues: Calibration, validation, robustness, social interface 887 Social networks and other interaction structures 917.1 Essential elements of SNA 937.2 Models for the generation of social networks 997.3 Other kinds of interaction structures 1047.4 Critical issues: Time and behavior 1068 An example of application 1098.1 The social dilemma 1108.1.1 The theory 1118.1.2 Evidence 1138.1.3 Our research agenda 1148.2 The original experiment 1148.3 Behavioral agents 1168.3.1 Fixed effects model 1168.3.2 Random coefficients model 1178.3.3 First differences model 1188.3.4 Ordered probit model with individual dummies 1198.3.5 Multilevel decision trees 1218.3.6 Classified heuristics 1268.4 Learning agents 1278.5 Interaction structures 1278.6 Results: Answers to a few research questions 1288.6.1 Are all models of agents capable of replicating the experiment? 1298.6.2 Was the experiment influenced by chance? 1318.6.3 Do economic incentives work? 1338.6.4 Why does increasing group size generate more cooperation? 1358.6.5 What happens with longer interaction? 1368.6.6 Does a realistic social network promote cooperation? 1378.7 Conclusions 138Appendix Technical guide to the example model 141A.1 The interface 142A.2 The code 145A.2.1 Variable declaration 146A.2.2 Simulation setup 152A.2.3 Running the simulation 157A.2.4 Decision-making 157A.2.5 Updating interaction structure and other variables 165References 173Index 187

Riccardo Boero, Economics and Market Analysis Team, Energy and Infrastructure Analysis Group, Los Alasmos National Laboratory, USA

Select Guide Rating

• Preface ix1 Introduction: Toward behavioral computational social science 11.1 Research strategies in CSS 21.2 Why behavioral CSS 31.3 Organization of the book 4Part i CONCEPTS AND METHODS 72 Explanation in computational social science 92.1 Concepts 102.1.1 Causality 102.1.2 Data 182.2 Methods 192.2.1 ABMs 192.2.2 Statistical mechanics, system dynamics, and cellular automata 222.3 Tools 252.4 Critical issues: Uncertainty, model communication 273 Observation and explanation in behavioral sciences 313.1 Concepts 323.2 Observation methods 353.2.1 Naturalistic observation and case studies 353.2.2 Surveys 363.2.3 Experiments and quasiexperiments 373.3 Tools 383.4 Critical issues: Induced responses, external validity, and replicability 404 Reasons for integration 434.1 The perspective of agent]based modelers 444.2 The perspective of behavioral social scientists 494.3 The perspective of social sciences in general 54Part iI BEHAVIORAL COMPUTATIONAL SOCIAL SCIENCE IN PRACTICE 575 Behavioral agents 595.1 Measurement scales of data 615.2 Model calibration 635.2.1 Single decision variable and simple decision function 635.2.2 Multiple decision variables and multilevel decision trees 655.3 Model classification 675.4 Critical issues: Validation, uncertainty modeling 706 Sophisticated agents 736.1 Common features of sophisticated agents 756.2 Cognitive processes 756.2.1 Reinforcement learning 766.2.2 Other models of bounded rationality 806.2.3 Nature]inspired algorithms 806.3 Cognitive structures 846.3.1 Middle]level structures 856.3.2 Rich cognitive models 866.4 Critical issues: Calibration, validation, robustness, social interface 887 Social networks and other interaction structures 917.1 Essential elements of SNA 937.2 Models for the generation of social networks 997.3 Other kinds of interaction structures 1047.4 Critical issues: Time and behavior 1068 An example of application 1098.1 The social dilemma 1108.1.1 The theory 1118.1.2 Evidence 1138.1.3 Our research agenda 1148.2 The original experiment 1148.3 Behavioral agents 1168.3.1 Fixed effects model 1168.3.2 Random coefficients model 1178.3.3 First differences model 1188.3.4 Ordered probit model with individual dummies 1198.3.5 Multilevel decision trees 1218.3.6 Classified heuristics 1268.4 Learning agents 1278.5 Interaction structures 1278.6 Results: Answers to a few research questions 1288.6.1 Are all models of agents capable of replicating the experiment? 1298.6.2 Was the experiment influenced by chance? 1318.6.3 Do economic incentives work? 1338.6.4 Why does increasing group size generate more cooperation? 1358.6.5 What happens with longer interaction? 1368.6.6 Does a realistic social network promote cooperation? 1378.7 Conclusions 138Appendix Technical guide to the example model 141A.1 The interface 142A.2 The code 145A.2.1 Variable declaration 146A.2.2 Simulation setup 152A.2.3 Running the simulation 157A.2.4 Decision-making 157A.2.5 Updating interaction structure and other variables 165References 173Index 187