Granularity (data)

Data scienceStatistics

Granular data is data that is in small pieces, for example in its most 'raw' form before being aggregated or summarized.

The advantage of granular data is that it can be molded in any way that the data scientist or analyst requires, just like granules of sand. Granular data can be aggregated and disaggregated to meet the needs of different situations. The opposite is aggregated data which is data grouped by particular categories.

Related terms

Data science

Anonymization

Data anonymization is a type of process to remove sensitive information to protect the identity of individuals in datasets.

Statistics

Confounding Variable

A confounding variable is an outside influence that changes the effect of the dependent and independent variable, potentially leading to false conclusions.

Statistics

Correlation does not imply causation

An easy mistake to make (a 'logical fallacy') is to draw the conclusion that, because two variables are correlated, one causes the other.