Poorly designed databases that allow for data redundancy will have data anomalies. A data anomoly is where an abnormal condition occurs because field value changes are required in many different locations. Therefore, a data anomaly develops when all the required changes in the redundant data are not made successfully. There are three types of anomaly:
Insertion
Data exists that should be added to the database but cannot be added as the data needed for the record is incomplete.
Modification.
Updating the value of a field involves changing the value of the field in more than one place. If all instances are not updated the data will be inconsistent.
Deletion.
Deleting data from a relation causes other data that is still valid to be deleted as well.