Continuous Data

Continuous data can have any value within a given range. Compare this to discrete data which is limited in the values it takes.

For example, the number of dots on a pair of dice or the number of wheels on a car limit you to a finite set of values. Measuring the size of the dice or the temperature of those wheels, with a precise enough measuring device, could give you infinite results. The dice might be 0.746″ and 0.748″, for example.

As a rule of thumb, continuous data is more valuable than discrete data. Now, in some cases, there is no option to gather continuous data. In the examples given earlier, the values shown on a pair of dice or the number of wheels on a car, both entail counting items.

In some cases, though, discrete data is a function of convenience. Think of the shoe size selection dropdown menu on a website. Even though the foot might require a size 10.39 shoe, that option is not available. You may see a similar situation when recording a person’s height into a medical database. It may limit the entry to every half inch.

These choices make sense. A shoe manufacturer would have to carry a great deal more inventory if they had to match infinite sizes. For the height example, there is little medical benefit to knowing how tall a person is down to the millimeter. A close approximation is enough for practical purposes, and it speeds data entry.

In other situations, though, the more precise the data is, the more you can do with it. For example, think of a quality issue. If you simply had a pass / fail metric (known as an attribute data type), you might know that 59% of an item was within spec. What you would not know, however, is any details about those that were out of spec. You would not know about the spread of the failures, if there were clusters of failures, or how far they were out of spec. With continuous data, you have much more information.

Keep in mind that if you collect the continuous data, you can turn it into attribute data. The reverse is not true.

Words of Warning About Using Continuous Data

• In practical use, continuous data is limited by the precision of the measuring device and the skill of its operator. Consider a tape measure with eighth inch tick marks. While measurements taken with it appear to be continuous data, in truth, the estimates between those indicators will likely be limited.
• Don’t overspend on collecting continuous data. Make sure the costs of collecting it match the value you get, or anticipate getting, from having that precise information.