Let us take an example model, spread of awareness model of Social Norms to describe the model and lets us deep dive in to see how the arguments that I have presented in the section I and II prevails. The models simulate the micro-behaviors of individuals about the consumption of a limited resource (Water or Energy). The chosen model aims to clarify the processes that leads a group of households to perceive a resource as ‘Critical’ for environmental sustainability and thereby to reduce its consumption and to observe the collective outcome of how a social norm emerges about the sustainability or unsustainability at a macro level (Sissa, G. (2014)). The objective of the model is to simulate the spread of awareness and assess the time required to reach the Reduction Goal in the system proposed. The models simulate, how awareness spreads in a community of agents, how the dynamic of such awareness impacts on individual reduction goals and on resource consumption, how the availability of smart metering functions can impact on such mechanisms. This agent-based model can be viewed as a helper system that assists in researching and theorizing a middle-range hypothesis of awareness spread in the given community. The underlying hypothesis is that ICT tools can enable and amplify key social and psychological mechanisms leading to environmentally sustainable lifestyles.
There is an overall reduction objective the system can reach or not. Attaining the objective corresponds to sustainable consumption or in short sustainability. The agents in the model are households. Agents don’t move and their position is always the same. This choice of non-mobile agents is driven by the consideration that agents are sharing the infrastructure where are available the smart metering functions, that are part of the infrastructure where households live. Such smart metering functions include: In home metering; Individual feedback about the individual own consumption of the limited resource; Information about green leaders and their low consumption profile that are taken as reference; Personalized advice for consumption reduction.
Awareness is a feature of each agent. It changes by interaction with neighbors in a given radius, by the influence of a green aptitude of a community and by a mechanism of social reinforcement. Agents are categorized into different types, and each type as different environmental awareness, different impacts on other agents as well as different awareness update coefficients. Because the categorization defines the consumption patterns and the potential reduction patterns, the awareness spread leads to behavior changes of agents in resource consumption. The agents are basically people involved in the consumption of one limited or critical resource. Each agent as mentioned earlier is a household.
There are mainly five types of agents; blinds, indifferent, spectators, actives, and evangelists. Blind Agents have negative environmental behavior. Their need to prevent overuse of the resource and their environmental sustainability goals are negative. Their consumption increases and they are mocking other green agents(i.e., actives or evangelists). Their awareness level is very low, and they have a significant negative influence on neighbors. They represent a block in achieving the tipping points. Normally, they don’t increase enough their awareness to change their type. They become more aware only if a large part of their neighbors consists of green agents and the social norms become evidently significant. They respond only to negative social reinforcement. Their consumption pattern is independent of the smart metering functions available. Indifferent agents are neutral about the environmental sustainability goal. Their consumptions are constants, with only some possible small reduction under very specific conditions, i.e. when they are supplied with a combination of smart metering functions. They don’t have an influence on neighbors but are influenced by them. Spectator agents are quite stable in their behavior but are open to listen and observe their neighbor’s behaviors. Under some combinations of smart metering functions, they can have reduction goal and They do not have an influence on their neighbors but are influenced by them. Active Agents are green people, engaged into a reduction of resource consumption. They have a significant positive influence on neighbors. They allow other people to look at their own data in order to show beneficial behavior results and to share reduction goal with others. They are responsive to positive social reinforcement. They are quasi-¬‐committed agents. Evangelist agents they are green activists that, in addition to active agents, can supply new resources into the system by producing the resource, for example when they produce renewable energy at a local scale with solar panels. They are prosumers. They have a strong influence on neighbors but are not influenced by them. Their awareness never decreases an evangelist. The agents know the global trend about resource consumption. When their behaviors are concordant with the general consumption trends the agents “reinforce” their beliefs and such social reinforcement in round change their awareness (Sissa, G. (2014).
Process- Macro- micro-level action and outcome
It is very obvious that the system tries to cover the macro- micro-level actions and their outcome in detail. In short, the process of the model is,
- Update of awareness: The awareness of the agents is updated according to the neighbors’ influence.
- Update of types: An agent changes his type when his awareness passes a given threshold.
- Update of reduction goals: The individual reduction goal varies according to agent type.
- Update of consumption: The own resource consumption depends on the reduction goal.
- Social Reinforcement: Social Reinforcement is a variable of each agent
During each run in the system, each agent looks around himself to verify how many neighbors and what type there are in the given radius. And according to these and other specific conditions he updates his awareness level. The awareness is also affected by a view of the overall “pro-environment” aptitude. After the update of awareness of each agent, when agent awareness is beyond a given value the model updates the agent type. Each agent has on own consumption pattern, this can be correlated to Merton’s view of micro action and the outcome. The pattern of consumption depends on the type of the agent and also on the availability of smart metering functions. Such smart metering function are the enablers to make agents able to measure the resource that he consumes, to have feedback on his individual consumption, to compare his own consumption with other agents, namely the agent with the lowest consumption. The process of agents getting influenced by their neighbors and their update of awareness all constitute to the micro action and outcome in the system. The overall change in the number of different type of agents and consumption of the resources in total by the population in the system there by attaining the overall reduction goal, constitutes the macro-level outcome of the system. This final macro-level outcome is reached through various mechanisms or processes such as Social norms, Social influence or reinforcement, and Smart Metering system. The change in the dynamic of the society is also captured in the model. Thus, the model allows us to implement a generative understanding of causality.
Association and Causality
It can be clearly seen that there is an association in terms of Resource consumption since the different agents in the system consumes resources differently and they all have different consumption rate. At the same time, no agents were affected with each other’s resource consumption or in other words no cause happened to an agent because of another agent. The result of the system such as, the change in global resource consumption, attaining sustainability or the unsustainability of the system, time taken to achieve the system reduction goal and the aggregate number of different types of agents can be causally related to the input criteria, such as initial setup of the system. Such as, the number of blind and active agents, smart metering facility in their households by defining mechanisms or processes such as social network and smart metering function.
We can choose Statistical and Qualitative cross- Validation for validating the paper and it can be done against the evidence that are qualitative in nature which describes about the behavior of individuals and how they interact within the agents. In the model, the consumption of resources is reported and the overall result of the process of decreasing the resource usage by awareness involving agents to inform and validate the model qualitatively at micro level. It can be perceived that this cross-validation of agent-based social simulation models is an achievement in analytical sociology. Micro validation against accounts of individual behavior and macro validation against the data that are aggregated can also be done. Reduction goal and individual resource consumption can be validated at the Individual level (Micro Validated). The overall resource uses can be validated at macro level.
The paper discussed agent-based models and why they are part of the empirical sociology framework in terms of building or contributing to middle-range theories and basic ideas on middle-range models, the definition of ABM, Causality and Mechanisms, and finally a brief overview of my experience in explaining the existing model and validating the model. It can be concluded that Analytical sociology provides the blueprint for understanding the social world and process. It brings out intricate connections between micro-level behaviors and macro-level patterns. Analytical Sociology assists in understanding a social phenomenon and predicting the outcome. The concept of middle-range theory can be considered as the bread and butter in Analytica sociology and it is the best method to define, analyze and study a social phenomenon. Middle-range theories are the best approach to explain a social behavior with the exact level of abstraction. To design and understand the Analytical sociology theories agent-based modeling system the best technique, as it clearly encapsulates the macro- micro-level of action and outcome clearly. Further, we have covered varying understanding of causality and how ABM fits into each in it. We can conclude that the ABM can be used for causal inferences as well, if it satisfies certain conditions. Moreover, all these was further argued and validated using an example ABM model, spread of awareness of social norms.