Data Type

Definition: A datum is categorized into one of many types that represent not only descriptive dimensions but also designate how the datum is to be used in analysis. Data types (not mutually exclusive) include continuous, discrete, binary, categorical, Boolean, date, time, memo, number, text, integer, nominal, ordinal, equal interval, etc.

What This Means: Knowing the data type is essential to understanding the proper analysis methods to apply to the data. For example, a continuous variable can be averaged, but a categorical/discrete one (binary such as yes/no, or on a five-point scale) cannot. That is why we revere standardized scores for assessments and warn users not to average percentiles. When defining the metadata for a collection, the data type is essential to ensure comparability, consistency, and conformance across all the data providers, repositories, and reports. As used in DataSpecs, data type is the data type of the item: Binary/BLOB, Boolean, Date/Time, Memo, Number, Text, etc. (The choices for this list can be edited by the System Administrator.)

Need another definition?