1- Can I do a survey using the DEMATEL (decision-making trial and evaluation laboratory) to establish a complex system illustrated as an influential-network-relationship-map, so, I can utilize that to build the CLD for dynamic balanced scorecard without doing interview?
2- After that, can I use the DANP (DEMATEL based on analytical network process) to confirm the influential weights, so, I can use weights to get the equations for building the stock and flow?
Dear Norah Ahmed Alsubait,
Among the MCDM and MADM methods, Dematel is one of the most suitable method could be mixed with SD. But it depends how to use it.
According to a real experience that I have in mixing these two methods, Dematel could be insightful in the qualitative step of System Dynamics modeling rather than the Stock-Flow model building. You could get some insights about the causality and polarity of that between two variables. For example, considering your research goal, it could help you to omit a very weak causality between to variable in your model. Also Determining the polarity between two variables based on Dematel method, needs a survey or interview with an acceptable group of experts.
More important than all: Remember that Dematel lacks a very important principle of SD and Systems Thinking, that is Feedback. So, If you want to use closed-loop feedback thinking, using the principles of System Dynamics and passing its steps is inevitable and Dematel is not enough by itself.
About your 2nd question:
You can confirm the causality and influential weight with any method but formulating a stock-flow model is more complicated than using those weights as formulas. Defining an equation between two variable in SD modeling should be based on Literature, theories, Experts knowledge and Systemic insight of the modeler. You may use the influential rates just for defining some coefficient in an more complicated equation.
You may take a look at these two works:
http://airccse.org/journal/jcsit/7215ijcsit03.pdf
Article A hybrid SD-DEMATEL approach to develop a delay model for co...
Best.
Thank you so much Milad
What do you think of using system dynamic first to build the CLD, then use DEMATEL to verify the model. After that I can use DANP to gain the weights (equations) for building the stock and flow
?
Note: my research is based on a dynamic balanced scorecard.
Hi again,
To reach a CLD, you could use both DEMATEL and main SD method simultaneously. DEMATEL could guide you regarding polarities and weight of causalities, But as i said it is not enough. You should develop dynamic hypothesis, using closed-feedback loop thinking an other systems thinking principles to capture the problem aspects. ( I will send you a reference book for SD modeling).
Then you wanna change the CLD into SFD (Stock-flow diagram) ; again using DANP could be useful but not enough. because it could only give you an influential weight , while mathematical equations are more complicated than just a coefficient of weight.Wights could be just some coefficient in a mathematical formula between two variable.
To know more take a look at the papers I send you.
Best.
Take a look at these documents:
Reference book of SD modeling:
Business Dynamics, Systems thinking and modeling for a complex world.
Papers:
https://www.emeraldinsight.com/doi/abs/10.1108/01409170810851276
https://www.researchgate.net/publication/228300662_The_Balanced_Scorecard_-_A_System_Dynamics_Modelling_Approach
Hi Milad
Please, I have more questions, you stated that ( According to a real experience that I have in mixing these two methods, Dematel could be insightful in the qualitative step of System Dynamics modelling rather than the Stock-Flow model building) and (Also Determining the polarity between two variables based on Dematel method, needs a survey or interview with an acceptable group of experts.)
based on what you have stated, this means:
1- I can use DEMATEL as a qualitative method by doing an interview with experts. is that right? however, I have read that the MCDM method is a quantitative method not a qualitative.
2- Can I use ANP as a qualitative method?
3- Can I use a hybrid DEMATEL and ANP (D-ANP) as a qualitative method as well?
4- If I used SD method to build my CLD, can I use D-ANP to confirm or validate my CLD?
See the attachment.
Yes, you cold use system dynamic first to build the CLD. See the paper:
BERNARDES, E. M. ; AMBRÓSIO, L. A. . DEMATEL Method as a Supporting Tool for Participatory Modeling. In: XI Congreso Latino Americano de Dinamica de Sistemas,, 2013, Ciudad de Mexico. XI Congreso Latino Americano de Dinamica de Sistemas, 2013. v. 1:45.
https://comunidadcolombianads.com/wp-content/uploads/2017/08/2013-Memorias-XI-CLADS-M%C3%A9xico.pdf
- Badges
- Science topic
- Similar topics
- Social Science