Categories: Error Measurement

Problem-Solving Methodologies

Solving problems is continuous especially in manufacturing/service processes to ensure integrity, efficiency, and thus profitability. Today’s solution methodologies would not have been possible without the input of Charles Bicking. A pioneering statistician, Charles Bicking the 1967 Shewhart Medal Award winner, provided an initial basis for methodologies of problem-solving by addressing six graphic methods (Bicking, 1983). His ideas of problem-solving methods were expanded on in 1984 and resulted in a teaching outline for learning statistical methods and techniques to enhance applicability to everyday challenging situations. The problem-solving methodology provided is used below; this is the basis for “statistical thinking.” To apply this way of thinking one must ask pivotal questions and find ways to visualize this way of thinking to garner answers to some simple and complex questions.

Questions are as follows:

1: What is going on in the activity?

The activity we are interested in could be the entire corporation, a business unit, department or just one process. The tool we will use to answer this question is the Flowchart. A flowchart is a diagram that help us to identify test sites/locations; sampling and feedback points; and redundant, missing, or obscure details of an activity. When one does not know the what, where, when, why, how, and who is doing an activity, then the problem-solver cannot begin to think about improving it. The analyst and decision-maker want to know what the process is and not what one think it is or could be. The solution lies in going to look at or examine the activity. This may include interviewing some people acquainted with the activity such as professionally researching authorities/experts on the subject matter. The best sources of information about a process are the people live and work in it, day after day.

2: What are the big problems?

Using the Pareto Diagram as the preferred tool one will be able to answer this question. This chart provides a means to evaluate how well we are/or have being doing in the past. With limited resources, and a multitude of potential areas to work on, the Pareto diagram provides us a means to prioritize our efforts. It can show us where the biggest problems in quality, cost, or quantity exist. It translates the data from an activity into a clear illustration of what we should tackle first

3: What are causes of the big problems?

Once it is known what the biggest problem is, there will be a need to determine what needs to be done to eliminate it. The problem may be due to numerous things (e.g., raw materials, equipment, poor or no training, unclear procedures, the inability to measure accurately and precisely something, or variability in temperature or humidity). It is best to use the Ishikawa Diagram to answer this question. Sometimes this tool is called a fishbone, feather, or cause and effect (C & E) diagram. The branches on the diagram graphically display the major factors and sub factors relating to the problem. Historically, the major causes were referred to as the four M’s, man, machine, methods, and materials. In the early 1980s, man was appropriately replaced with persons or people. I have found it beneficial to add a branch to the diagram for environment. With this tool, often using brainstorming techniques, we probe the relationship of possible causes of our problem.

4: What does a look at the past data show us?

After one select a probable cause to the problem, there will be a need to examine historical data for analysis. For this it is preferable to use the Histogram to answer such a question. Knowing different kinds of data (measurable, count, proportional, etc.) form particular shapes or frequency distributions is a great help. The histogram enables one to determine whether the data fits the theory for developing a solution and whether some unusual, unexpected, and/or undesirable behavior is present that should be eliminated. For example, if our data are measurable (inches, pounds, feet, etc.), then the expectation is for the bell-shaped “normal” distribution. However, if a Poisson shape should be seen, then the process must stop, and a determination made as to why the shape is not what was anticipated. Unexpected shapes are often the result of unknowingly using nonhomogeneous data, such as multiple machines or shifts, unfit or outdated parts from different machines or suppliers. The histogram reduces the chance of making an error in the subsequent steps of our analysis.

5: Is the hypothesized cause and effect relationship real?

Astute time management skills are needed if control is to establish whether changes to a selected “cause” will create an impact on the effect/action one will want to change, improve, or eliminate. For this, the tool one can use to answer this question is the scatter diagram. When one concludes a relationship does not exist, then there will be a need to go back to Step 3 (Question 3) and select the next most likely cause.

6: What does the current data show us about this activity?

The Control Chart can be used to answer this question. The control chart provides a means to monitor an activity (process) while it is running. This tool can alert us to abnormalities (changes) and outliners in the process so corrective actions can be taken before they become chronic problems. These abnormalities might be due violations of procedures or failure to ensure the integrity of adjustments a person or machine made through changes in materials, manufacturing process or changes in people, such as a shift change. As a preventive tool, it also tells us when things are running the best according to specifications/expectations. Therefore, the control chart demonstrates when to do plan an activity, and when to leave such activity as is.

The incomparable Dr. W. Edwards Deming (1985) once said, “If we go about it right solving problems may be easier than we think.” These questions are aimed to establish thinking processes using statistical methodologies and techniques to take mitigate problems before they happen.