Accuracy. Does the data describe the reality as it is? Accurate data describes the object
represented in the database with the right values, both in form and format.
A typical mistake regarding format accuracy is recording dates with wrong format. If the
database format is MM/DD/YYYY and someone registers dates with the DD/MM/YYYY format, the
data stored in the database won’t reflect reality correctly.
Uniqueness. Is the data recorded with no redundancies? Avoid unnecessary duplication of
information as much as possible. As an efficiency principle, each piece of data should be
recorded once, which means that the number of objects in the real world matches the
number of entities represented in the database.
If the database registers victims of violent events by capturing biographical and socioeconomic
characteristics, each victim should be recorded once. If the same victim suffers two different
violent events and gets registered twice, there will be a duplicity. To ensure uniqueness, each
entity —in this example, the victim— is identified with a primary key that can be related with
various violent events.
Timeliness. Is the data available when it is required? How long does it take from when an
event is noticed until its data is properly registered? How much time passess from data
collection to database recording?
Completeness. Are all the required data items properly recorded? How many blank values
are there in the database? Blank values are empty slots in the database, which means that
there is no value for a given property. A high percentage of blanks in the same category
might be a signal of a broader issue, as misunderstanding of its meaning or need of
reformulation.
If the database is supposed to help to provide telephonic assistance to victims but the telephone
of a victim is not recorded, the database will fail in its mission due to lack of data.
Consistency. Are there any contradictions between data? Does the data match across
different data stores?
If the database registers the victim’s date of birth and, also, the victim’s age within a range of
ages, there should be no contradictions between both fields.
Validity. Is the data valid according to the indexing rules? Databases often include syntax
rules to set the data format, typology and range. Some common data types are text,
5