I think many people are unaware of the fact that data analysis is designed to establish a broad consensus and that consensus is supposed to reflect an approximation of reality over time.
That means firstly that everyone must be reminded that they are modeling reality, not representing reality. There are assumptions about the model which need to be kept in mind. If these assumptions are not appropriately understood and well communicated, there is likely to be discord in the ‘now what?’ phase of decision making.
This also means that assumptions may change, and reality may change. So new models must be made. Older ones should not necessarily be discarded, but remembered as heresies against the new regime. There is always value in studying heresies because one cannot always validate assumptions and certain assumptions may revert back over time.
When people are focused on charts and graphics and technologies without paying attention to the process and social dynamics of decision making, they are only seeing a fraction of the problem they are trying to solve. These are the reasons why aphorisms exist like:
"There are three kinds of lies: lies, damned lies, and statistics."
The most difficult and vexing questions to answer are:
1. What were we thinking when we made that decision?
2. Who knew what and when did they know it?
3. Why were we not paying attention to X?
These questions always show up in the wake of a disaster, and the inability for people to address them shows the incompleteness of their thinking around the matter of modeling a changing reality, establishing consensus, and keeping track of the appropriateness of assumptions.
In other words, governance of the decision making process is very valuable and often overlooked (or assumed to be automatic) when a system of analysis is successfully put in place.