How much rigor is enough?

[Gene Bellinger]

After reading Jim Lyneis' '08 presidential address and fabricating an entry for WikiSD, http://www.systemdynamics.org/wiki/inde ... m_Dynamics, I found my elf asking, "How much rigor is appropriate?", which lead to this story, which doesn't start with "Once upon a time."

Arbor Networking has been growing steadily for almost a year now and the support organization has been continually struggling to hire and train enough people to maintain the level of support readiness to server their customers. The rest of the organzation is growing also, and continually looking for people that are familiar with Arbor's products and clients. And, where do you think is the most likely place for the organziation to find these people; support of course. Which makes the sitation even worse for support.

I've attached a graphic which depicts this situation.

Just as the organzation is about to produce a policy statement that prohibits the rest of the organization from hiring people from support some enlightened indivdiual looks at the sitatuion and says - duh? Seems like a good pause statement. The policy statement would be good for support, though make things worse for the rest of the organization. Though by considering the structure in it's entirety it is realized that it's pretty much a Boulding "Prey/Pretor" system. As it turns out the real enligtened response to this situation is to provide support with additional budget and staff and make developing people that are attractive to the rest of the organzation part of their mission -- and the whole organzation benefits. It's a lot like just creating more food for the rabbits so there will be more of them for the foxes to eat.

So, in conclusion, my answer to my own question is: "One should never apply more rigor than is adequte to deal with the sitaution." So now I just have to ponder how one gets a handle on what's adequate.

Thoughts?

be well,
Gene

[Geoff McDonnell]

In the design sciences (ilke engineering, medicine, management) there is always a tradeoff between rigor and relevance. Therefore we need enough rigor to ensure the model is fit for purpose. If the purpose of SD models is to produce effective action, then we must target our rigor (and the way we present the dynamics) to the audience we wish to convince to act effectively.