Community vs. Social Network

Hello  Bob,

Thanks for the quick turn around and the discussion.

I will always try to respond if I can. If I don't it's simply because I just don't have enough time. I rarely give up on interesting discussions, because ever since I left academic, I miss these nerdy chats... I'm actually sick and resting at home now, so I have some time today.  ;-)

I'm glad to hear that you have your own framework that works for you. I totally get what you are saying. Communication is a very valid perspective. I just like to look at community vs social network at an even more fundamental level, more stable and longer term, because the way we communicate changes rather quickly. But if that suites your need, that is great! It is a good model for what you are trying to do.

I tend to be a purist when It comes to Occam's razor, and adopt the simplest model that explains the most observation. But as I said before, since social sciences are not rigorous, there is no way for any model to be complete. It comes down to choosing between the following scenarios. Suppose there are only 3 phenomena (say, A, B and C) that we care to explain.

   Model 1: explains A 60% and B 70% and C 65%

vs.

   Model 2: that explains A 90%, but 10% of B and 10% of C.

   Model 3: that explains A 10%, but 90% of B and 10% of C.

   Model 4: that explains A 10%, but 10% of B and 90% of C.

Most social scientist tend to create models that explain their data (observations, phenomena) without consideration of all the other data out there. So they tend to create many different models (theories), that explain their observations very well. I, on the other hand, would rather sacrifice the explanatory power for individual phenomenon to get a more holistic understanding of all three phenomena. Otherwise, I would have to adopt many different models. But if phenomena A is what you care about, then clearly Model 2 is better. But if you need to explain only B, then Model 3 will do just fine. Likewise, if you don't care about A and B, then Model 4 is probably better.

See my point? It's a matter of taste, which you pick. I just like the elegance and simplicity better. Moreover, there is more internal consistencies. Since it's just one model (even though it's not as good as other models with respect to different problems), I don't have to worry about reconciling the difference between the models.

Anyway, that is the trade-off I constantly have to make when constructing my models and perspective. They are not necessarily the most impeccable models, but they tend to be the simpler and more elegant ones. 

Thank you for your validation and I appreciate the discussion.

And I'm glad to hear that you are reading the book. It's written more for a business audience, so it's definitely not as rigorous as I like to be. But let me know what you think.

Thanks again for the conversation and hope to see you next time.