Good engineering combines data- and fact-driven analysis with experience, hunches, and a sense of what's going on.
I have a long list of complaints about the style and approach of today's so-called, alleged journalists. For many of them, actual fact-based reporting of what is happening is the last thing on their minds, since they would rather "tell a story" and "put a face to the news". But I'll focus right now on one technique that I see way too much of, and is an easy trap for all of us to fall into: extrapolating a trend or defining a situation based on very shaky foundation.
Specifically, I saw this recent article in The New York Times, "Digital Devices Deprive Brain of Needed Downtime", here. In order to humanize the story, it started with several paragraphs using a named, individual as an example. Of course, this is completely meaningless, since you could just as easily find a different but contrary example.
It then attempts to generalize from this person's situation, with a string of anecdotes and quotes, but no data at all. I've heard the term "anecdata" for this approach of using of a lot of anecdotes and isolated factoids to attempt to create the aura of a credible case, when you have no actual data you can cite.
As icing on the cake, the story includes a vague reference to some unspecified study using mice. Here's the dirty little secret: you can always find a mouse study, somewhere, to support just about any contention, that's what graduate students are for. (Did you know that mice fed a healthier diet prefer classical music over rap music? I am sure there is a study showing that, or maybe the opposite! ;-)
Ironically, I happen to agree with the premise of this article, but that's irrelevant here. What I don’t like, and I worry about, is when a few scattered quotes and anecdotes are used to create a scientific-sounding premise and then substantiate it.
Why does this concern us? When we are doing project planning, design, and debug, it's important to separate out data-driven analysis, troubleshooting, and conclusions from those which are driven by anecdotes, hunches, and our gut sense of what is going on. Please note: there's absolutely nothing wrong with the latter group, not at all; in fact, those "soft attributes" are the result of experience and often the key to a projects completion and success.
But you can become so enamored of these softer attributes that you neglect to look long and hard at what solid data is available and then reconcile the data with your feelings. Don’t dismiss one or the other, but work to understand where they agree and why, and just as vital, where they differ and why that might be so.
We're fortunate that we have sophisticated, accurate software and hardware tools and test equipment that can help us, and we are obligated to use it as part of our projects. But we also have the human brain, which works in very mysterious ways to figure things out, to see old and new patterns, and to make leaps of reasoning. We shouldn’t slight either one, but we should admit and weigh the role and place of each in the bigger picture. ?