Podcast with Lynda Finn: The Value of the Simple Run Chartby John Hunter
This podcast is the first episode of our “Knowledge In Variation Series.” Lynda discusses the importance of moving from spreadsheets to plotting data, and the common mistakes that organizations make if they aren’t charting their data.
She starts by sharing some of the hardest things for people to grasp about the Deming philosophy. Though it varies, Lynda finds it’s most difficult when Deming’s ideas don’t align with the practices people feel have contributed their success.
Having the context and seeing what the process is trying to tell you about where it is headed by looking at the data in run chart order I think is, in general, 90% of the benefit of a control chart or more complicated method of looking at patterns over time.
So yes, that is what I am advocating just a run chart. Just plot the data over time. If the number is worth being on the report it is worth seeing in its proper context. So lets get that context out there in front of people.
Linda discusses why all of us should look at data in a chart showing the data plotted over time. Along with a few simple rules on what makes a result a signal of a special cause (meaning the effective problem-solving or process-improvement strategy is to examine that special result closely can guide organization to more effective action. Largely this is by guiding people to temper their tendency to react to many results as special causes.
Looking closely at a specific result is helpful if the result is “special”; has an identifiable special cause for that result. But in most cases the result is do to common causes of the system and the most effective strategy to improve is to look at all the data and seek to find improvement to the overall system that will change the overall performance of the system.
Acting on data in a spreadsheet or 2 or 3 points of data (last month and this month and maybe last year) results in a great deal of waste looking for special causes of those result that are just the natural result of the current system and variation. By looking at data over time and applying a few simple rules we can properly understand the data as signaling no special cause and therefore signaling we should look at all the data to seek solutions to improve results.