8127/frame/fm

 
g.
Validation
Validation studies should be in place for procedures that the laboratory
performs on a day-to-day basis. The internal reviewer should verify that the
analyst’s interpretation of the results falls within the limitations of the data
and is supported by the validation data. For new techniques, or substantive
adaptations of established techniques, the reviewer should ascertain that
sufficient validation has taken place to support the reliability of the test for
the purpose used in the case.
The independent reviewer will also be interested in whether sufficient
validation has taken place. Sometimes this can be accomplished by reviewing
a literature of published articles. However, for new or obscure methods,
published studies may not be available. In these cases, the reviewer must
depend on studies performed by the laboratory, the field, or increasingly, a
commercial supplier of reagents or equipment. In these cases, he should
attempt to obtain this information through discovery. This is not always
possible. Although laboratories are becoming increasingly open about mak-
ing this information available to independent experts, the privilege has also
been abused by certain of those experts, causing some laboratories to estab-
lish strict policy restrictions regarding access. An additional hurdle concerns
commercial companies who claim proprietary rights over certain patented
elements of a product, or refuse to disclose internal studies validating its
8127/frame/ch09 Page 239 Friday, July 21, 2000 11:44 AM

---

240
Principles and Practice of Criminalistics
reliability. This practice seriously undermines the scientific basis of the testing
procedure. Without full information, the scientist is limited in investigating
the cause of anomalous results and must rely on the company to troubleshoot
the procedure. Extensive empirical validation by the client laboratory can
counter these concerns to some extent, but does not substitute for complete
disclosure.