Page 201 - Automated Fingerprint Identification Systems (AFIS)

186 AUTOMATED FINGERPRINT IDENTIFICATION SYSTEMS

identification of the same individual, it would be counted as one hit under
option 1, one under option 2, but four under option 3.
The calculation of the number of idents becomes even more complex when
the method used to calculate a hit rate is considered. The survey asked the
following:

Which of the following is used to determine a hit rate:

1. All latent cases received by the site/latent unit (regardless of whether they were
searched on an AFIS system).
2. Only Unique cases searched on AFIS systems regardless of number of times
searched on AFIS systems (one case received and entered on an AFIS system
regardless of the number of searches performed equals one case).
3. Multiple searches of the same case (one case received, entered on an AFIS system,
and searched five times equals five cases).

Three users choose option 1 and calculate their hit rate on the number of all
latent cases received regardless of whether they are searched on an AFIS system.
Nine users use option 2 and calculate their hit rate on the number of unique
cases entered ignoring any additional searches of the same case. Three users
selected option 3 and based their hit rate to include multiple searches of the
same case. (See Tables 8.2 and 8.3.) What are the consequences of these choices?
Some agencies use a very narrow definition, counting only one ident regardless
of the number of lifts hit, while others take a more broad approach. Also, some
agencies include all latent cases in their ident rate regardless of whether they
were searched on an AFIS system. What is the effect of different latent entry
practices, different databases, and different ident hit rate calculations?
What are some of the factors beyond count that may contribute to a higher
ident rate?

• Better control of what is collected at the crime scene
• An evidence technician collected the latent image at the crime scene
• Search parameters were optimized
• The best card was selected for conversion
• Database size and volatility
• Accuracy of the matchers
• Accuracy of the coders

As a practical application, Table 8.4 provides a comparison of the responses
from two agencies, one a local police department, the second a state agency.