8127/frame/fm

 
Standard
: A measurement system against which an unknown sample is
compared. It can be numerical, generating a quantitative
result, or comparative, giving a qualitative result. An example
of a numerical standard would be a simple ruler. An example
of a qualitative standard would an element with which an
unknown substance is compared.
 
Control
:
A sample whose result is known. This assures the user that the
test system is functioning as expected. An example of a positive
* The ASCLAD-LAB report resulting from the inspection of one particular laboratory
decried the absence of a “positive control” for footwear analysis. This is one of those
instances where a no doubt well-meaning inspector applied a general quality criterion
without thinking about it very hard. Perhaps a vision test of the analyst would show that
he could see within proscribed parameters on the day the comparison was done.
8127/frame/ch09 Page 229 Friday, July 21, 2000 11:44 AM

---

230
Principles and Practice of Criminalistics
control is a sample containing a known percentage of alcohol
that is used to calibrate a breathalyzer. A negative control con-
tains all the reagents in a test system but no sample.*
When we are training new analysts, or worse yet, old supervisors and
managers, the question inevitably arises regarding when it might be acceptable
to omit the proper controls and standards. Gee, my runs are only 5 minutes
apart; I just have to rerun this one sample that I diluted improperly. Gee my
runs are only 1 hour apart; I haven’t changed any of my reagents between this
run and the last one; my runs are only a day apart…. Well, you get the idea —
when do you draw the line? Good laboratory practice is like the religion of
science; the philosophical generalization always wins over any situation in
which one might like to manipulate an exception. It is not up to us to figure
out ways to circumvent good laboratory practice. Like the “do it right or do
it again” dilemma, one soon learns that it is not worth trying to eliminate a
sample or two, because this attitude will inevitably backfire. The run in which
it turns out you have a marginal evidence sample will be exactly the one in
which you have decided it’s just too much trouble to include the calibration
controls that assure you that the instrument is operating at peak sensitivity.
Quality assurance guidelines and quality control procedures are neither
merely an encumbrance nor a panacea. When implemented thoughtfully and
with intelligence, they are simply another tool used to demonstrate the high
quality of a work product.