The Power of Observation, Part 1

Charles Kettering, the famous inventor, once said: “A problem well stated is a problem half solved.” This implies that a good portion of problem solving should be devoted to a thorough understanding of what’s going on before any corrective action steps are taken.

In many cases, too much time is spent on proposing various solutions before the problem has been correctly defined. Observation is a powerful technique that can be used to help understand problems. Often, clues about how to solve a problem can come from simply observing the process. Following are a couple of examples.

One particular problem I was involved with concerned damaged product being received by an internal customer. A team had worked on the problem previously and had made the assumption that the root cause was due to shipping damage. They took measures to improve the packaging and better protect the product during shipment to the customer. This resulted in adding cost, but didn't protect the customer who continued to find defective product.

Once involved, I tried to determine the origin of the problem. Analyzing the customer feedback, I found that two second-shift operators accounted for 48 percent of the problem. My first step was to simply observe these two operators to determine what, if anything, they were doing differently. Almost immediately I saw what was causing the problem. The operators were literally dropping the product into the shipping container and damaging it before it left the manufacturing area.

On another occasion, I was asked to help improve a process that was generating excess scrap. The process was manual in nature and the defects generated were very obvious. I lined up several of the defective units side by side with some good parts to visually observe any differences between the defective and good parts. I noticed a nonrandom pattern in the defective parts, which was the result of an operator using a different assembly method. Once this was determined, standardized work was generated, and all operators were trained to the proper method.

Observing is using one or more of your senses—sight, hearing, smell, taste, and touch—to gather evidence or data. You want the observations you make to be accurate and objective and avoid opinions and bias that are based on specific points of view.

Qualitative and quantitative observations

There are two types of observations: qualitative and quantitative. Qualitative observations are descriptions gathered as a result of your senses and do not use numbers. Examples are “the sky is a light blue” or “it is raining very hard.”

Quantitative observations, on the other hand, are descriptions based on measurements or counts and do include numbers. If you count objects or measure them with instruments, you are making quantitative observations. Examples are “there are 10 people at the meeting” or “the nonconformance is 15 mm long.”

Consider the following tips for making observations:

• Use all of your senses (sight, hearing, touch, and smell) to make qualitative observations.
• Review your observations to be sure they are accurate and objective.
• Whenever possible, count or use measurement devices to make quantitative observations. Be sure to include units with your measurements.

Check your observations to be sure that they are statements about information gained through your senses, not explanations of what you observed.

Next, let’s talk about inferences. An inference is an explanation of what has been observed. When you make an inference about observations you have made, you are logically explaining what your observations may mean. You can make many inferences about the same set of observations. The key is that any inference must be reasonable and logical. For example:

• Reasonable inference. Makes sense, given what you know.
• Unreasonable inference. Assumes too much from the evidence.

Inferences must be checked through investigations to find out if they are correct. Here are some tips for making an inference.

• Base your inference on accurate qualitative or quantitative observations.
• Combine your observations with knowledge or experience to make an inference.
• Try to make more than one logical inference from the same observation.
• Evaluate the inferences. Decide what new information you need to have so that you can prove your inferences are true. If necessary, gather more information.
• Be prepared to change, reject, or revise your inferences.

Observation has the advantage of not disrupting the current operation of the process, and any clues that exist are not disturbed. In their haste to make improvements, too often teams start by twisting knobs and making changes before they really understand what’s going on. This usually has several unwanted effects. First, the dominant cause of the problem may get masked and delay the team from finding it. Second, band-aid solutions get implemented and add cost that may not be necessary.

Taiichi Ohno, the architect of the Toyota Production System, had a teaching method of placing a student inside a circle located in the production area. The student was instructed to stay inside the circle and observe what was going on and look for ways to improve it, continually asking the 5-Whys. They would look for ways to improve product flow, prevent the seven forms of waste, and make the process more efficient.

A team improvement tool based on this same method is the Kokai watch, as Gerald F. Smith describes in his book Quality Problem Solving (Quality Press, 1998). During a Kokai watch, a team observes a department for a period of time without asking the operators any questions. They look for ways to improve performance related to quality, efficiency, and safety. These improvements may be better information availability, better material handling methods, better work place organization, etc. After observing the process, team members discuss their findings and recommend corrective actions. This technique is very powerful and only requires that the process be observable. Workers should be assured that they won't be reprimanded and that improvements are only aimed at the system.

In the second part of this article [1], I’ll discuss some of the better observation tools and techniques that will help you better understand your organization’s processes.

Note: This article was first published in AIAG's May 2009 Quality, Standards and Tools newsletter.