Evidence supports that content of what is discussed — while words are important — can almost be ignored and focus on the pattern of speaking. A group in which one person talks all the time don't reach their shared goals together; whereas more effective ways of interacting are short bursts of information from everyone, and lots of different perspectives are represented, so the group can reach the common goal more effectively.
This area of study (documented in a book of the same name) is not as focused on the digital spaces — rather high-bandwidth face to face spaces that we are evolved to work in. What they created measures the volume of people and the proximity of people speaking to each other. They get offices and families to wear this to collect a lot of data. There is two experiments...
The first one is almost an excuse to build an openbadge prototype, get this made (it's open hardware) and try it out — there is a hub (github.com/HumanDynamics/openbadge-hub-py) to visualize behavior that may reflect how people e.g. work in a Hackathon.
The goal is to give people perspective of an overview of speech patterns, as well as exploring new forms of idea in group dynamics.