Without a doubt it is chal­leng­ing to pro­tect and serve our com­mu­nity.  I write a lot on this blog about the short­com­ings and lim­i­ta­tions in foren­sic sci­ence today.  Be assured that there is a lot of good sci­ence in foren­sic science.

No mat­ter how far we come, in some are­nas the per­sis­tence of the myth of valid­ity per­me­ates.  One such area is the con­tin­ued use of canines at road­side.  Canines are used at road­side to try to develop prob­a­ble cause to search a vehi­cle or used at other places to try to detect illicit drugs or explosives.

Leash-lead on a drug dog

In one recent CNN news arti­cle Truck sus­pected to hold explo­sives found full of energy drinks the fol­low­ing is documented:

A police dog’s false hit for explo­sives on a truck that had been dri­ven errat­i­cally by a man who did not speak Eng­lish attracted a throng of law enforce­ment per­son­nel to a rural road in Wis­con­sin… But they turned up noth­ing more pow­er­ful than energy drinks

But this is not unique.  In fact 60 min­utes did a full seg­ment on this topic in 2004: Does The Nose Know?  Some Drug And Bomb-Sniffing Dogs Are Unre­li­able. Within the story it remarks on the Rus­sell Lee Eber­sole affair where he was con­victed on 25 counts of fraud after pro­vid­ing the State Depart­ment, the IRS and the Fed­eral Reserve with bomb-sniffing dogs that, when given a test, couldn’t detect 50 pounds of dyna­mite or 15 pounds of C-4 plas­tic explosives.

Auburn Uni­ver­sity pro­fes­sor Larry Myers, one of the country’s lead­ing experts on canine detec­tion pro­grams in that pro­gram summed it up well when he said:

They can tell you that something’s there, that’s not there, sim­ply to get praise, to get food, to get what­ever they’re work­ing for.  It is a tremen­dous prob­lem. We have train­ers that can’t train. Dogs are being used that can’t –don’t know how to do anything.

The war­rant­less non-consensual intru­sion into the prop­erty of another by the gov­ern­ment with­out prob­a­ble cause should not be excused by a dog­gie sniff.  It just is not right.

Each dog is unique.  No dog has been truly sci­en­tif­i­cally val­i­dated in terms of its pur­ported sen­si­tiv­i­ties or selec­tive­ness.  While it is per­haps impos­si­ble to do so, it does not mean that we should accept that the dog is selec­tive, spe­cific sen­si­tive and its results valid and true.

Some of the fre­quent prob­lems sur­round the issue of cuing the dog.  Dogs, like humans, are mas­ters of body lan­guage.  Even well-trained dogs have a par­tic­u­lar sus­cep­ti­bil­ity to pick up even sub­con­scious cues from their han­dlers.  This is why there is a need to not leash-lead the dog.  Also the han­dler should also wear sunglasses.

It is impor­tant to remem­ber that there are basi­cally two types of trained dogs:  active alert and pas­sive alert.  It is impor­tant to know the lim­i­ta­tions on both.  It is also impor­tant to note that dogs are not robots.  They have feel­ings, wants and desires.  They want to feel spe­cial.  They want to be rewarded.  Some­times they will alert even against their train­ing in order to get the reward.  Some­times, they will alert sim­ply to stop hav­ing to work due to their envi­ron­ment, such as it is too hot or the dog is tired.

In addi­tion, most dogs are not dis­trac­tion trained.  At road­side or in the field there is a lot of dis­trac­tions that can over­whelm even a well-0trianed dog and lead that dog to alert so it can be removed from the over­load of the envi­ron­men­tal stimuli.

It is crit­i­cal for the Prac­ti­tioner to get the dog’s alert record, which should be kept by the han­dler.  This record can pro­vide a mean­ing­ful source of cross-examination in terms of apply­ing sta­tis­tics such as like­li­hood ratios and per­cent false positives.

It is most impor­tant to remem­ber that dog­gie sniffs are not val­i­dated science.