How do you see the IMPAC II database? Each of us sees IMPAC II in a different way, but there is only one database. Any changes we make may affect data retrieval for everyone. Nowhere is that more obvious than in the area of personal data.

In the IMPAC I system, personal information was record-specific. IMPAC I record-specific personal information is the same as IMPAC II Role information. An individual has a different Role ID for each committee or grant, and each Role ID can have different information. Following Role IDs through time, for example, might show the same person with a first name of William and no middle initial, or a first name of Bill and no middle initial, or a first name of William and middle initial "C". In the IMPAC I system, there is no field that identifies the same individual across records.

IMPAC II adds a Profile ID, a record that does not exist in the IMPAC I system, to identify the same individual across records. The Profile ID has the same fields as the Role ID and contains a summary of the information on the Role records. The Profile is created with the first incoming Role record, so the data it contains may be incomplete. William's Profile, since it can only have one name, may have been created with a first name of William and no middle initial. Some information (degree codes for example) may have multiple values. If the Profile ID was created with a PHD and the next incoming record had an MD, the MD would be added to the Profile. Now William has a dual degree, but only on the Profile.

In general, data from a Role will be added to a Profile if the field isn't already filled (for example, a SSN will be added to the Profile if the SSN field is empty) or if the data can have multiple entries (degree codes). Data does not usually flow in the opposite direction (for example, training Roles should not acquire degrees through time). It is important to understand that many Role IDs (from training, committee, and grant data) are linked to their single Profile ID. All Institutes will use the same Profile ID for a person (there is only one) even though each Institute has its own unique Role IDs for the person.

Those individuals who have worked with the Consolidated Grant Applicant File (CGAF) and are familiar with its set number will see a similarity between the set number and the Profile ID. Both are intended to identify multiple instances of a single individual. Although the CGAF assigned a set number, it did not create a new record with corrected, summarized data. IMPAC II attempts to improve upon the CGAF by creating a single Profile record that contains current and complete personal data, maintained by authorized users.

To better understand the Role/Profile construction in IMPAC II, it is important to examine the relationship between the data in the OLTP and in the IRDB. In the OLTP, the identification number associated with a record is the Role ID. As the data flows from the OLTP to the IRDB, however, data linkages change. An individual's single Profile identification number replaces unique Role identification numbers. In William's case, the OLTP will report two grants, each with a different name format and a single degree. The IRDB will report two grants, with the same name and a dual degree.

Accurate Profile information is critical to retrieve accurate personal information from the IRDB. If the Profile information is incorrect (for example, if someone incorrectly changes a name or a degree code), all IRDB records associated with the individual will have identical, and unfortunately incorrect, information.

Users who change records hold the key to personal data quality. In IMPAC II, users may correct, or inadvertently corrupt, an individual's personal data. The ability to change records includes the responsibility to understand the consequences of changes to Profile and Role records. Changes to a Profile record affect not just the user's data, but also potentially all Institutes' data.