"Just because something happens after something else happens doesn't mean it happens because something else happened
So, what we have here is the tyranny of the "Three Cs" *- Show me your data
- Show me your assumptions
- Tell me what questions you asked
- Tell me why you didn't ask other questions"
From the TED Radio Hour, "Big Data"
- Coincidence, perhaps better written as co-incident to emphasize the timing of multiple incidents, but having no other coefficients or linkages among them
- Correlation, meaning one effect or outcome is predictable upon the occurrence of another, though the "other" outcome may have many contributors, so the correlated effects may be "weak"
- Cause-and-effect, meaning one effect or outcome is both predictable by- and, indeed, caused by the presence -- or not-presence -- of another
It seems we always are confronting the confusing rule that "correlation is not causation, but causation requires correlation"
Chapter 2 of my book "Quantitative Methods in Project Management" goes into these ideas in more depth (did I mention it's available at any online book seller?)
Of course, when there is a human in the loop, then all of the correlating or causative linkages are influenced by biases, most non-linear and very situational, but some amenable to "game theory" which is discussed in other posts in this blog
* There is a 4th C: co-variance, an ideas from statistics. Related to correlation, co-variance describes how the spread -- or distribution -- of uncertain or random outcomes of one thing is made different in some sense by the presence of another random or uncertain outcome.
If there is zero co-variance, then the uncertainties are "independent" of each other; otherwise they are not.
In most project environments we can't actually measure co-variance; what we can do is test for independence and thus infer a co-variant effect, or not.
Read in the library at Square Peg Consulting about these books I've written
Buy them at any online book retailer!
Read my contribution to the Flashblog
