Introduction
Let’s begin with a couple of definitions:
- Causality = the NECESSARY relationship between incident (cause) and result (effect), which is a direct consequence of the first.
- Correlation = the measurable association between two variables or events.
Causality is, by definition, empirical, whereas correlation is, by convention, anecdotal.
Consider the following statement:
A flood happens one evening when a comet is visible in the sky; therefore, the comet caused the flood.
The two events (the appearance of the comet and the flood) are CORRELATED but not causal. They happened proximally in time (contemporaneously), but any further attempts to link the two beyond that (without further information) is futile, as there is no evidence that the flood was a direct consequence of the comet’s presence, rendering the statement above logical fallacy.
Now, consider this statement:
Students who eat breakfast do better in school.
Is this causal or correlative? It certainly seems causal, right? But, the causal interpretation would state:
Students that do well in school do so BECAUSE they eat breakfast.
Looking at the information from this viewpoint suddenly makes the statement seem silly. Just because a person eats breakfast may have some impact on his/her scholastic performance, but it is hardly the CAUSE.
Finally, let’s consider:
Smoking causes lung cancer.
Here is a true causality. How do we know that it is not just another case of correlation masquerading as more profound information? Because, as with any EMPIRICAL evidence, the information was gathered in a controlled study (a study in which a test case is compared with a case that has known effects or outcomes), and is therefore consistent, observable, and reproducible.
How Does This Apply to Process?
At the heart of agility (the process as well as the physical characteristeric) lies accuracy, of which speed is a by-product. If we are to become agile as a team in our response to emergencies as well as planned, controlled work, we must improve not the rapidity with which we respond, but rather the accuracy with which our responses are targeted.
In any scientific setting, empirical evidence reigns supreme, though often difficult to obtain. Often, in less formal scientific settings, we are romanced by the allure of evidence that APPEARS to be empirical, but with careful consideration it rapidly dissolves under scrutiny. Our agility will be predicated by our ability to focus on the empirical and recognize the anecdotal.
I cannot tell you how many times in my development career I have looked at a piece of code that is under-performing, and KNOWN, beyond a shadow of a doubt, what the problem is, only to stick a profiling tool on the code and be proven wrong. Often, the area of the code that was the culprit was a place I never would have looked.
The point of this is even though we have an intelligent, skilled team with a good deal of experience, we need proper process and tools to increase our accuracy. Jumping to conclusions based on insufficient, vague, or downright incorrect information doubles the amount of work required to solve the problem. Stepping back, examining the facts, and isolation and removing variables will ensure we are seeing a clear, proper picture of the issue, so that we can target responses to it with surety that they will move us toward a solution of the core cause rather than a band-aid on a symptom.