Hypothesis Based Problem Solving

Problem-solving using a hypothesis-based approach

Thinking back to my days of doing puzzles and quizzes, I remembered doing mazes. Just how did I get to the prize in the middle? Well, I always cheated. I started at the end and worked backwards, and it was still much more comfortable.

The same can be said of problem-solving. Think about the endpoint (solution) and work back from there. That is the principle of a hypothesis-based approach.

What is the hypothesis?

Hypothesis – an idea suggested as a possible explanation for a particular situation or condition but has not yet been proved correct – source Collins Dictionary.

What is problem-solving using a hypothesis-based approach?

Firstly we need to define with precision what the problem/opportunity is?

Write it down and gain agreement that this is the issue.

To help clarify, use the 5 Whys tool and check that the problem is the problem (industrialist Sakichi Toyoda developed the 5 Whys). It ensures that you understand the root cause, not just the symptoms, i.e. what you can ‘see’ (particularly useful if you have a recurring problem). At each ‘Why’, you will identify potential courses of action. However, a thorny issue may require a more complex approach.

It is a simple technique often used by 5-year-olds when they don’t like the answer you have just given them…

Develop an action plan

What are we going to do? Who is going to do it? When will it be completed? Are there any dependencies? What do we already know? What don’t we know?

Analyse the data available. Use the data to test our hypothesis. Does this back up our belief?

Are the gaps significant? Do we need to buy/commission more data? Run trials?

Be strict, ask yourself, “what question am I trying to answer?”

Update the action plan and go back around the loop based on the findings or impact on timings. Confirm or amend hypotheses.

Using a hypothesis-based approach to problem-solving should deliver results quicker and with a better return. It is an efficient way to validate/create solutions, reduce time spent on aimless analysis, and more time on research to deliver results.

What are the potential explanations? What are our hypotheses?

Don’t re-invent the wheel. The answer may already be ‘known’, i.e. a hypothesis. Gather a group together and work through the potential solutions to solve the problem / deliver the opportunity. Throw out any that are not plausible based on knowledge /experience.

You can use a logic/issue tree to capture the parts.

Prioritise. How big is each solution (if relevant)? How will it contribute to solving the problem? What will it cost, and how long will it take? Identify the answers you want to take forward, break them down into parts so that actions/responsibility can be divided and allocated. Remember the 80:20 rule too. Identify some easy wins and some more complicated steps.

Share