"Good problem solving" shows up on most job descriptions, recruitment messaging and performance appraisals. Thinking about problem solving in such generalized terms doesn't communicate much.
There are different types of problem. Competent problem solving is not the same across problem varieties. Some are simple: clear, well understood, single presentation with agreed solutions. Some are chaotic: no one knows what's really going on and there is no consensus on how to proceed. See our earlier post on appropriate approaches in these first two problem types: http://www.currycorp.net/blog.html?id=182
Even more challenging and dangerous problems populate the other two quadrants. The third group are problems with actively pursued consensus solutions despite being poorly understood. Let's call those ‘crazy mad problems' because of the frustration produced when predictably poor or negative results occur. Then there is the fourth group, problems that are well understood but have no agreed upon, non-contradictory solutions. Those are ‘wicked problems' because adequate resolutions require more than knowledge and logic.
Crazy Mad Quadrant 2 is our focus in this post. These are problems where our knowledge base in insufficient and yet there is wide consensus on suggested action. A hundred years ago, this was blood-letting for treatment of gout and purgatives for depression. Fifty years ago, this was completely immobilizing fractured limbs. Twenty-five years ago, this was bed-rest for muscle-strain injuries. Fifteen years ago, this was hormone-replacement therapy for all menopausal women. Five years ago, this was shaking off a concussion.
We convince ourselves that the state of our knowledge is sufficient to reduce the problem to the simple variety: if X, then Y. This ill-founded self-assurance is bolstered by wide agreement on recommended action. We do not notice or bother with carefully following up results, especially unexpected results. Poor outcomes are shrugged off rather than being correctly perceived as counter-factuals.
Good problem solving in this quadrant begins with providers or originators careful, documented application of the agreed protocols combined with extended critical attention to results obtained: positive and negative, expected and unexpected outcomes. Problem solving continues with reflection on the compiled data at both the point of origination and beyond. Routine self-reflection is a hallmark of professionalism, no less so in circumstances that we have convinced ourselves are simple and straightforward. We can't learn that the problem is not simple unless we look at the poor results honestly. This is the value of routine audit. Increasingly the audit function is conceived as an administrative function separated from provider or originator self-reflection.
Making negative results known is essential in this quadrant but has been difficult to accomplish. There are well-documented difficulties in getting negative results published in academic journals. In reaction, there is a movement to support routine data repositories and public access to both the original data and to the compiled, audited and critically appraised evidence base.
So, what percentage of what you think are simple problems are really the crazy mad variety? Have a critical look at your success rates, and how that success has been defined. Blood-letting didn't usually kill patients immediately.
We will look at the wicked variety of problems in the next post.