O'PEEP'S DISCRETE DATA EXPLANATION
You ask yourself whether a piece of data is discrete or continuous? Easy! If it feels like climbing steps - it’s discrete data. You can only put your foot on a defined step, or a piece of data on a defined level. If it feels like walking up a normal hiking trail it’s continuous data. You can put your foot anywhere. Same applies to the data - it does not have to be on defined levels, it can be anywhere.
Discrete Data is a numerical type of data that can only take certain values and is usually determined by counting.
Data can be discrete:
due to the nature of what's being studied
(e.g., number of workers. There can not be 4.5 workers.),
because the source of the data is attributive
(e.g., number of good and bad parts).
Discrete data is usually displayed using descriptors like proportions or visualized using pie and bar charts.
From an information and statistical standpoint, it is always better to use continuous data rather than discrete. With continuous data we have more information, we can detect variation and statistical calculations are more reliable and powerful. Moreover, with continuous data you typically need less individual data to make a sound conclusion of what's being investigated.
The delivery performance of two vendors (A and B) is evaluated. First in a discrete manner, only providing good/bad information, and then with continuous data according to the delivery time in minutes and the specification limit.
As you can see, continuous data provides us with more information because it not only tells us if it was a good or bad service but also how good or bad it was. While just looking at four data points it is difficult to conclude that A is always better on the left side (discrete data) while on the right side (continuous data) we have more confidence to say A is better with the same amount of data.
Join our next Green Belt Training to get certified in systematic process improvement.