How to start making a System Dynamic? "Before starting to play with the dynamics, we try to identify the system first, also the SYSTEM LIMITATIONS. so we know where we are going to start and the scope of the study so that it will not continue to expand the subject matter to be studied. "
Things to Consider in Creating Models Dynamics
Tips System Dynamics
- How to start making a System Dynamics?
"Before starting to play with dynamics, we try to KNOW the SYSTEM first, also the SYSTEM LIMITATIONS, so we know where we started and the scope of the study." - "In recognizing a simplified system with causal loops, then when Stock Flow Diagrams, the most important thing to consider is the sorting of the variables to be entered, which ones are Stock / Level, flow and auxiliary and constants (It should be sorted / grouped first). Do not get confused, between auxiliary and level (Stock, Flow and Constants), so that it can be easier to build the structure of the system later"
Tools System Dynamics
- "Understand the advantages and disadvantages of the tools that will be used" Examples of tools for implementing dynamic systems: Powersim, Vensim, Stella, Ithink Because the most important thing is philosophy and system thinking. Visual Basic, Borland etc. can be used as additional tools to help create a display interface (because it is a programming language, so it can make any software).
In recognizing the system, before making a causal loop, the most important thing to consider is the sorting of the variables to be entered. Which is stock / level, flow and auxiliary and constants (Should be sorted in advance / grouped). Do not get confused, between auxiliary, level and others, so that it can make it easier to build the structure of the system later.
"LET'S START WITH A SYSTEM APPROACH" Things to note are:
- Determine the objectives of the system to be reviewed
- Determine the variables that will be seen behavior
- Make a needs analysis from the field / group / sector / stakeholder, etc., so that later you can see the combination of the dominant level of needs
Sorting requirements analysis based on criteria, to facilitate the development of problem formulations. Ex. Stakeholders (Government, Community, Business Actors, NGOs and Academics); Dimensions (Social, Economic, Environmental, Technology, Law and Institutional). Needs analysis can be developed more broadly or narrowed down, depending on the study to be conducted.
Focus the problem so that it is in line with the general problem (in the list of all existing problems) from all stakeholders, fields and other aspects, as well the problem is able to lead to the objectives to be achieved. In the formulation of the problem can be obtained based on the results of previous studies, with analytical tools or with expert judgment.
In identifying the system, several stages that need to be considered include:
To make it easier to see the relationship between variables in the system should be described in the form of a causal loop diagram. Make a causal loop from a common model first, then it can be divided according to the needs of each sub-model.
To make it easier to see the relationship between variables in the system should be described in the form of a causal loop diagram. Make a causal loop from a common model first, then it can be divided according to the needs of each sub-model.
Be careful not to reverse making a Causal Loop relationship. The safest way is to see from the formula / equation that we made. But to make it easier to make the condition of the cause variable is a real or fixed state, we can only know the positive / negative of the Causal Loop.
From the causal loop diagram (CLD) that has been built, list and sort (using tables) all constants, auxiliaries, flow and levels including units, whether there are variables in the CLD or outside the CLD but can affect and be influenced by the system, don't forget to include influential factors even as "given" to the model.
On the part of building the structure of the system and input data it should be little by little "running", so that it is easy to create and see the interrelationships between the variables drawn in the formula.
pay attention to the behavior of the system on the variable being simulated (or all output variables), avoiding negative values. Generally there is no negative value, although logically there is (ex. Money / savings if negative means debt, it is better to make a new variable as debt).
The data used in the model should be time series, it is used for the validation calculation process, which aims to assess the objectivity of a scientific work, because objective scientific knowledge must obey facts. Except the newly developed model, so there is no time series data, such as social studies related to attitudes, perceptions and behavior.
The data used in the model should be time series, it is used for the validation calculation process, which aims to assess the objectivity of a scientific work, because objective scientific knowledge must obey facts. Except the newly developed model, so there is no time series data, such as social studies related to attitudes, perceptions and behavior.
Validation Test Can Be Done
- Structure Validation, by looking at logic and units in the model being built
- Output / Performance Validation, by looking at the graph difference between the simulation graph and the real time series data graph. Done in validation <10%
- Field Validation, To ensure simulation results with field conditions. Mainly related to land use simulations
Use the factors that have been selected (problem formulation), leverage or dominant factors to simulate policy scenario. Because of factors those selected in the focus of the policy making study. If the selected factor is qualitative quantification should be done either from the results of scoring or the results of experts.
