Page 200 - Automated Fingerprint Identification Systems (AFIS)

STANDARDS AND INTEROPERABILITY 185

presented by the author at the 1999 Educational Conference of the Interna-
tional Association for Identification.
Fifteen AFIS managers were asked to complete a questionnaire about their
identification practices and policies. The questions ranged from how latent
prints were captured at crime scenes to how they were counted at the conclu-
sion of an AFIS search. From the answers provided some conclusions can be
drawn. Almost immediately it became apparent that the managers participat-
ing in the survey represented two major user groups: police departments and
multi-jurisdictional state agencies. The police departments collectively had far
more investigative personnel than the state agencies, and had more direct influ-
ence on the evidence collection process. In general, the law enforcement agen-
cies collected the latent prints and either forwarded them to the state system
or used the state AFIS system for their latent print searches.
Law enforcement managers are familiar with the practices of their depart-
ments and how they interact with the state AFIS system. They may not, however,
be as familiar with the technical aspects of the AFIS system as the state admin-
istrators are. For example, it is not necessary to know the size of the AFIS data-
base to use an AFIS system as a search tool any more than it is necessary to
understand the bank interchange process to use a debit card in Paris.
The responding states do not all have the same responsibility with regard to
criminal investigation and may have a slightly different focus. State AFIS users
have provided information as it applies on a state level. For example, database
size, entry practices, hit rates are provided for the entire state AFIS network,
not just the practices employed by a single agency.
The areas of most interest were how the idents are counted and how the hit
rate is determined. There was no uniformity in the answers. The survey asked
the following question:

How do you count your latent identifications?

1. One ident per case, regardless of number of lifts hit.
2. Multiple idents per case, only if more than one individual is identified with lifts
from that one case.
3. Number of latent prints hit (i.e., three lifts hit in one case from the same SID
produces three latent idents).

Six respondents chose option 1, four, option 2, and five, option 3. The method
chosen has a serious impact on how many idents are counted per case. For
example, if four different individuals are identified in one case, those idents
would be reported as one hit under option 1, four under option 2, and four
under option 3. In another example, if four lifts in one case result in the