reading review - problem solving 101 (riff offs, part 1)

Problem Solving 101 by Ken Watanabe (September 2018)

Problem Solving 101 is a simple, short, and highly popular introduction to the basic process of problem solving. Watanabe’s target audience is business bros kids and his book is filled with the basic examples, cute graphics, and neat conclusions that appeal to business bros kids. Despite being the definition of the target audience outside the target audience, I enjoyed the book and would recommend it to anyone who is interested in a quick refresher on the subject.

I always have a difficult time doing full reading reviews for books like this one. I suspect there are two reasons. First, the book is short and I therefore took only a few notes. Second, the book focuses heavily on examples and this led me to write my notes in a generalized way that allows them to stand on their own without being put in the context of the book.

So, instead of the usual reading review I’ll just riff off the ideas I noted down. Unlike in the past, though, I took a moment to rank these thoughts in reverse order of my preference. We’ll do a few today and come back with the rest later in the week.

Good luck, reader.

Tim

10. The steps of problem solving: (i) understand the situation, (ii) identify the root cause, (iii) develop an action plan, and (iv) execute until the problem is solved, making modifications as needed.

Like I mentioned, this book is very simple…

This note highlights an important feature of good instructional books – they assume nothing about the reader. The result is a very thorough dissection of a skill most people probably already feel they do fairly well. Readers who come into this type of book expecting to learn something new on every page misunderstand the purpose. Rather, what a book like Problem Solving 101 does by its thoroughness is help a reader pick out one or two details to tweak in order to achieve better performance. Most readers will consider this book a series of reminders rather than a mind-blowing instruction manual.

9. Problem solving is a habit that combines certain skills with the right attitude.

This is the thesis statement for the book. From the list in note #10 above, the skills involve (ii) and (iii) while the attitude covers (i) and (iv). To put it another way, the right skills mean knowing how to analyze a root cause and develop a plan while the right attitude means asking questions until you understand the situation and confirming the problem is solved after execution.

8. Asking a series of yes/no questions can help create a list of possible ways to solve a problem. With the right series of questions, it is possible to place every option into a certain category.

This is a good example of the book’s preference for thoroughness. Going through every possibility until you’ve listed all the possible options isn’t necessary for solving every problem, of course, but it is probably the most important thing to know how to do for solving the most difficult problems. I suggest practicing the approach whenever possible so you can be ready to employ the method when you need it.

Most people who informally think of ways to solve a problem call it ‘brainstorming’. This works just fine for most problems. I would recommend simply ‘brainstorming’ whenever everyone involved has a good sense of all the available options. However, Watanabe’s approach of asking yes/no questions is recommended when there are unknowns about the choices because the process of asking the questions forces decision makers to clarify the details involved in the decision.

7. Collecting information and performing analysis just for its own sake is a common trap. Make sure such activity is done to answer a specific question or to achieve a specific goal.

This is the most important thought in the book. We live at a time when information is more easily accessible than at any other point in human history. The temptation is to take all the data, look at it with an unbiased point of view, and ‘let the data do the talking’ (1). Unfortunately, conclusions derived in this manner often fail the strict validations demanded by the field of mathematics or statistics.

Instead, I recommend asking a specific question and considering how information might help answer the question before performing the analysis. This approach might require more work. However, the benefit of the approach is that you will be able to accept the results just as they are and safely move on to the next step of the analysis.

6. A hypothesis must be tested to know the magnitude of a given input on a result. A given input might improve an outcome by 10% but alternate methods could achieve much more.

I’ll try to clarify this sloppy note with an example. Suppose you commute the same way to work every day and that your total time is thirty minutes. You think this is the fastest way. But how do you know for sure? The answer is to form a hypothesis then test it. Maybe you cross the street at the same intersection every day – why not cross at a different intersection for a week? If your commute for the week drops down to twenty-five minutes as a result, well, you might be on to something there, reader.

The point of the note is that testing a series of hypotheses can help clarify what is really going on in almost any situation. The process of properly testing means determining how you would prove or disprove the hypothesis before conducting an experiment under unbiased conditions (2). This process is helpful not just for problem solving but also for ongoing process improvement because it requires the same kind of mentality – if my hypothesis were true, how would I know?

Footnotes / a 24 reference? / hypothesis testing 101

1. But what if the data knew where a bomb was about to go off in an hour?

A metaphor I’ve heard more than once regarding analysis of ‘big data’ compares the process to torture – you just keep poking and prodding and putting your cigarette out on the data until it tells you the truth. My response is always – how come torture is the best comparison we can come up with? Shouldn't we be suspicious of any result derived from torture?

2. One more thought on hypotheses…

Another note I took down brings additional clarity to the ideas I discussed in point #6 above:

It’s helpful to know if your hypothesis is taking on a grouping structure or an argument structure. In a grouping structure, components of the argument are independent of each other and refuting one does not necessarily crush the argument. In an argument structure, each pieces builds on a neighbor and refuting one can break the entire chain.

This thought further emphasizes the important of understanding exactly how to refute the hypothesis before starting the experiment – if this step is ignored, you run the risk of misinterpreting the results.