Models as Tools to Assist Development of Shared Understanding

Models are the means by which humans translate perceptions into information, knowledge and institutions [56], and they are the workhorses of modern science and management activities. When we think of models, we most commonly think of quantitative predictive models. However, models have multiple roles and functions: model development can assist researchers in understanding the structure and processes that contribute to emergent system behaviors; models can help define an envelope of possibilities, a range of possible outcomes, for selected policy options and future conditions; models can provide insight into the sensitivity of the system to particular actions and changes, thus allowing managers to make the most effective use of scarce funds and staff time; and models can serve as a means to communicate and integrate diverse viewpoints.

Selection of policy and of management actions is based on "If-Then" hypotheses, whether explicit or implicit. These hypotheses are derived from models: sometimes the models are quantitative and precise; more often we unconsciously rely on our mental models. Any model is a simplification of a complex reality, and thus is inherently incomplete and approximate. Such simplification is a fundamental element in the function of the human brain, which filters an effectively infinite amount of data in the form of sensory input and puts the most significant bits together to form a mental model, or schema, that approximates the momentarily significant attributes and behavior of the real world. The critical question for evaluating a particular model in a particular situation, whether a mental model or a quantitative predictive model is whether or not it offers a useful simplification. When developing an explicit model intended to guide policy development and collective behavior, a consensus on simplifying assumptions is required in order for people to trust the model, and to accept any decisions based on the model and/or the modeling process. Strategic filtering of information—identifying what can be left out of the model—involves judgment calls based in large part on tacit knowledge and creative insights. Confidence in the result of the filtering process comes from direct involvement of end-users and stakeholders in the filtering process.

In many management situations, reasonableness and believability are more important criteria than predictive precision for model selection and development. There are inevitable trade-offs between accuracy and precision in modeling, and increasing model detail will not necessarily result in more accurate predictions or reduce the risk of making a bad prediction. Walters discusses three reasons for distrusting detailed models as much or more than simple ones [42]. First, critical interactions or events can be highly concentrated in space and time at scales, locations, or times that have been ignored, or over which a simple averaging process has incorrectly been assumed. Second, adding more detail adds more parameters to the model structure, and it is probable that there will be inadequate data available for at least some of these added parameters. Finally, because of feedbacks and cross-scale interactions, we seldom have accurate enough data on process rates and initial spatial pattern to accurately simulate the outcomes of events [57]. As a result, models are rarely fully predictive, and that they are best thought of as illustrative [11].

For example, a cause-effect relationship identified by biophysical scientists is that an increase in the proportion of impervious surface in a watershed will result in increased levels of pollutants in stream water and an increase in extreme water flows, both high and low. This relationship is relatively easy to express as a quantitative model which can be used to predict the approximate reduction in surface water pollution that will result from removal of a given amount of impervious surface. Effective policy aimed at improving surface water quality and water flow based on this relatively straightforward relationship is not simple to formulate, however, and the quantitative model provides little guidance. The amount of impervious surface in a watershed is a consequence of decisions made by a large number of people and organizations. Rather than a single cause-effect relationship and a single if-then hypothesis, there are a number of linked relationships and hypotheses that can be thought of as a causal chain. Some links in the chain are purely social in nature, some are purely biophysical, and some connect the two. A more comprehensive model is needed by policy makers, one that includes information about the driving forces behind the choices that people make related to impervious surfaces. The inclusion of human behavior means that such a model cannot be quantitative and fully predictive, but it can provide policy makers with important insights that can assist them in crafting effective policies.

Management actions based on an invalid or inaccurate model are likely to produce unexpected and undesired outcomes. The effectiveness of a model in capturing a particular aspect of dynamic system behavior depends on the appropriateness of the assumptions used to filter and prioritize the flood of potentially relevant information. Because every complex SES is unique, reliance on the general knowledge of experts over the specific experiential knowledge of local peoples is likely to produce an inappropriate (or, at least, non-optimal) set of assumptions, particularly those related to human behavior, and the resulting model is likely to be a poor predictor of system behavior. In other words, development of useful and accurate models of SES behavior requires the participation of people with diverse experiential knowledge of the specific SES of interest. Given the extreme complexity of SESs, everyone can be considered "non-expert," since even technical and scientific experts will be expert in only part of the whole system. Knowledge about a given SES's structure and behavior is distributed among a large number of organizations and individuals. Therefore, it is critical to include a diverse group of people, with a wide range of expertise and experiential knowledge, in the model development process. Finding effective ways of integrating data has proven to be a continuing headache, however, because of the volume and disparate nature of the information generated by the problem-solving process [30].

Collaborative development of a conceptual system model is an effective starting point for building both knowledge of what to do and the collective capacity to act. Walker and colleagues suggest that development of a conceptual model of an SES based strongly on stakeholder inputs can serve to bound the problem and to elicit information on the important issues, the major drivers, and to identify key areas of uncertainty [8]. A conceptual system model can assist in recognition of cross-scale interactions and human scale biases.

Model development should an iterative and long-term process. A model is a snapshot, a summary statement of what is understood about a particular system at a particular time. The system will change over time. In addition, our collective knowledge of a given system will always be incomplete, and so our understanding of the system structure and dynamics will evolve through experience and new scientific findings. Updating the conceptual model can be an effective means of integrating new knowledge into ongoing and future decision processes.

Negotiating Essentials

Negotiating Essentials

Always wanted to get a better deal but didn't have the needed negotiation skills? Here are some of the best negotiation theories. The ability to negotiate is a skill which everyone should have. With the ability to negotiate you can take charge of your life, your finances and your destiny. If you feel that others are simply born with the skill to negotiate, you should know that everyone can learn this wonderful skill.

Get My Free Ebook

Post a comment