The Detailed View, the 30,000 Foot View and Compromising Between the Two
As engineers, we are often caught between the nitty-gritty of the data we generate and the grand vision we see for the future. There can be a tug of war between those two perspectives, as if one must choose between counting the trees and seeing the whole forest. What is the relationship between these two views? How do you balance both levels of detail to achieve the overall objective?
Granulated or Ultra-Detailed View
• You need granular data to validate a hypothesis.
• Detailed information collected over time is necessary to identify root causes.
• Detailed data analysis turns generalized information of root causes into equations that describe the direct relationship between an input and an output or an event and its effect.
The Relationship Between Detailed Data and the 30,000 Foot View
• Granulated data creates the graphs on which overall trends are identified, measured or countered.
• Detailed data is necessary to generate general estimates of the impact of different process changes.
• Most of your audience, unless they are fellow engineers or accountants, related better to the big pictures with some additional information than long discussions of statistics and charts.
Balancing the 30,000 Foot View with Detailed Data View
• Always collect granular data before accepting common knowledge that there is a trend, relationship or tendency.
• Recognize that your data will generate outliers, even in six sigma quality processes.
• Know that exceptions do not invalidate generalizations, unless the exceptions are so common that Statistical Process Control rules would cause you to question underlying assumptions. To quote Dennis Prager, you cannot act as if a generalization itself is wrong because exceptions exist, only that the generalization itself is wrong. Exceptions don’t disprove a generalization, because it’s a generalization. And that is the 30,000 foot view your internal and external customers see.