Before actively looking for solutions, you should ensure that you understand the problem and its root causes. That is, diagnose before prescribing. Our four-step process makes this explicit—and it makes implicit sense to do so, too—and yet, we constantly fail to do so.
Consider British Midland Airways Boeing 737. On January 8, 1989, it was cruising at 28,000 ft when the flight crew sensed a strong vibration. Fumes and a burned smell led them to believe that one of the engines was malfunctioning. When the captain throttled back the right engine, the vibration stopped, which led the crew to believe that the right engine was the problem. So they turned it off.
However, the loss of vibration was coincidental, and the left engine was the one malfunctioning. In the ensuing activity as they tried to reach their diversion airport, the crew did not validate the nature of the problem. The left engine eventually failed completely during their final approach, and the plane crashed with neither engine running. The accident killed over 40 passengers [1, 2].
Dealing with complex problems, we deal with evidence that is often inconclusive, contradictory, and messy in other ways. For instance, in general, an item of evidence is compatible with several hypotheses, which may make it easy to reach the wrong conclusion.
Focusing on the why (as in, why does the problem exist?) before the how (as in, in which different ways can I solve it?) can help us avoid some of these mistakes. That is, especially when it is not critical to solve the problem quickly, we should take the time to understand its root causes before looking for solutions.
1. Orasanu, J. (2010). Flight crew decision-making. Crew resource management. B. G. Kanki, R. L. Helmreich and J. Anca: 147–180. [p. 160]
2. Air Accidents Investigations Branch (1990). Report on the accident to Boeing 737-400 G-OBME near Kegworth, Leicestershire on 8 January, 1989 (Aircraft Accident Report 4/90). HMSO, London.