1) The modeler seeks guidance on the development of a model in the area of the examination of policy in natural resource management
Helen is studying the likelihood that market-based mechanisms, specifically the offsetting of emissions with carbon sinks, will support improved natural resource management, specifically the reduction of dry land salinity, in the Western Australian wheat belt. She has provided drafts of a short paper and PowerPoint presentation for her presentation at the PhD colloquium at the conference site on Sunday just prior to the start of the conference on Monday. Her PowerPoint presentation presents several preliminary causal loop diagrams and simulation model runs, and contains the following modeling questions:
1a) "How do I determine which are the driving variables?"
1b) "Scale - Ho do I deal with hierarchies?"
1c) "How do I deal with qualitative values and attitudes? When there are 2 or more qualitative values, both 0.5, 0.5 * 0.5 = 0.25?"
1d) "How do I deal with the aggregation issues - data rich/data poor?"
2) Adding more structure to the model for his paper entitled, "Acceptable Risk and Mitigation Options: Dynamic Structure of Building Safety" David writes:
"I have developed a conserved flow model to study policies that determine the distribution of building safety in communities. I’m presenting a paper on this model in one of the sessions at Palermo. After the meetings, I want to expand or complement this model by 1) adding revenues so that the costs associated with various policies can be taken into account, and 2) adding some structure (or creating a companion model) to show how the goal of overall community safety gets displaced. I’m an experienced social scientist, but a novice in system dynamics. It would be great to talk through what I’m trying to accomplish with a modeling expert."
Here is the abstract for David's paper from the preliminary conference schedule:
"'This paper reports the results of a study designed to understand and facilitate disaster mitigation in communities characterized by low frequency/high magnitude earthquakes. A system dynamics model describes the distribution of building safety over time and the dynamic structure governing the decision-making that created the current distribution of building safety in a small town located near the New Madrid Fault Zone and therefore at significant earthquake risk. Data from this town is used to establish a 20-year baseline. Simulations are run over a 40-year period to examine the consequences of different building policies and the effects of a magnitude 7.0-earthquake in the year 2002."
3) Development of a model for educating business people about the benefits of spending money on safety, health, and accident-prevention. Antonio writes:
"I'm trying to carry out a SD model concerning the prevention against accidents at work in a firm (above all SME, but not only). The aim is especially educational: my Institute (a public body, see below) will use this model during free-of-charge training to show the entrepreneurs (or small businessmen) the economic benefit coming out of cost for safety, health and accidents-prevention at workplaces."
"At the moment, the model is only in my mind. I'm looking for ideas and I'm trying to start a task force to deal with accidents-prevention problems through System Dynamics. I have been working with SD and making models for six-seven years."
4) Relating discrete time maps and system dynamics models to one another. Paul writes:
When introducing system dynamics (SD) to some people, I'm sometimes asked to translate to system dynamics a discrete time model with which they are familiar. In attempting to do this I am often confronted with two problems. First the units of the discrete time model often do not seem to be internally consistent. Second, the behaviors of the two models are often quite different. I suspect the error is somewhere in my translation, and of course, these problems lead to difficulties in explaining SD. This is discussed briefly in footnote 14 on page 290 of Sterman's Business Dynamics in the context of the logistic map, with references to Gilbert Low's treatment of Samuelson's multiplier-accelerator model. To overcome my translation problem, I'd like help converting these two discrete time models to system dynamics, with a discussion of units in the discrete time models and equivalency of behavior of the two models. Specifically,
1) Sterman's footnote refers to Gilbert Low's 1980 paper "The multiplier-accelerator model of business cycles interpreted from a system dynamics perspective" as an example. Yet the equations and sketch in Figure 4.1 of Low's paper don't seem to match. What should the equations be to match the multiplier-accelerator sketch?
2) Returning to Sterman's footnote, he writes,
"Every discrete time system can be converted into an equivalent continuous
time system..." What is the meaning of the word
"equivalent" here? Shouldn't the behavior be generally the same for both the
discrete time and continuous time systems (the discrete time map can produce
chaotic behavior, but I don't think the system dynamics logistic model can
produce chaotic behavior)? Or is the
"equivalence" only structural, not behavioral? To
address this question, I would like help converting the discrete time map
x(t+1)=kx(t)(1-x(t)) to a continuous feedback model with internally consistent
units, and then comparing the
behaviors of the two models.
5) Framing/formulating a model of student traffic on a university campus, and exploring the
unintended consequences of building a new 'precinct' on the campus.
"On the particular campus, it is proposed to build a new commercial hub adjacent to a car park in what is now remote from the existing
commercial centre of the campus. The sponsor of this model has significant concern
about this initiative and the unintended consequences. It is
hypothesised that there will be a shift in traffic patterns which will then manifest itself in decreasing the financial viability of existing
commercial activities."
6) Modeling knowledge creation using system dynamics.
Soebagijo is studying knowledge creation in the cellular telecommunications corporations in Indonesia. He writes, "I am now composing a knowledge creation model in the 'usual model' (the relationship of various variables), and wish to convert it to a system dynamics model. I am also considering 'learning capabilities' as another alternative focus of interest." Sebagijo will have his model available in Palermo (it may be posted here before Palermo).
7) Finding data to test your model
Elise is presenting a paper, co-authored by George Richardson, entitled "'Threshold Setting and The Cycling of a Decision Threshold." The abstract for her paper from the preliminary conference schedule is:
"'When policy makers use a test result with a cutoff score in a decision, the cutoff threshold may change over time. An example is the threshold of "reasonable suspicion" used to justify a police search. Hammond (1996) postulated that a decision threshold will oscillate over time in response to competing pressures from affected constituencies, as unavoidable cases of false positives (e.g. innocent people searched) and false negatives (e.g. guilty people overlooked) emerge from the uncertainty of using an imperfect test (e.g. level of evidence) to predict the actual measure of interest (e.g. guilt). The structural underpinnings of a cycling threshold are analyzed in this theory-building article. First, we present a simplified converging model of Hammond's initial insight. Then, we present three alternative models: one with integral control representing the historical dissatisfaction of competing constituencies; a second model with delays in policy maker responsiveness; and a third with stakeholders shifting constituencies."
Elise has the following questions:
"I'm curious about the use of data in modeling -- what if you have an idea but are not clear on where to find data to test it? I am wondering about how to apply my work that I will present in the parallel session to a real-world situation. My question relates to possible applications, and how to move from theory to application." More generally, " how do SD practitioners make client contacts and what is the process of model development in a real-world context?"
8) Bass Diffusion Model in Undeveloped Countries
João writes, "I would appreciate some assistance on the usage of the Bass Diffusion model in underdeveloped countries, where income inequality and demographic growth play a important role. Trying to cope with these factors, my colleagues from CPqD and I have added some extra constructions to the basic Bass model. I would be glad to have this modified model reviewed by a coach and, if time permits, to discuss other aspects of the Bass model (such as "p" and "q" as time functions). I will be presenting, on Tuesday afternoon, the poster #697 ("Dealing with Complexity in Telecom: A System Dynamic Approach for Planning New Services and Technologies")"
Here is João's abstract from the preliminary program:
"'Telecom markets are undergoing rapid change all over the world. Economy globalisation, deregulation, technology evolution and increased competition all present continuous challenges to telecom players. In order to successfully operate in this new scenario, telecom operators and service providers must use new approaches, tools and techniques to face the demand for new services and applications by users. The complexity and dynamism exhibited by such a scenario and by the features and technologies inherent to new telecom services, make a system thinking/system dynamics approach especially suited for the job. This paper presents the methodology for telecom service planning that has been used by our team at CPqD. The methodology incorporates a systems thinking/system dynamics approach in order to support business modelling and the risk analysis. The paper also shows how the methodology is put to work, helping a software developer to evaluate and compare different business models associated with an electronic courier service."