That is a interesting piece. While, we do not do "Scent lineups" and I do not know of any US Police Agencies doing them, they have been done in Europe. There is no doubt in my mind that a dog can do it. I train with a SAR person that does this and it is very impressive. Working a dog in several disciplines limits what we can do, but I can always aspire to be better. Training with SAR folks that put so much time, effort and emphasis on trailing has made me and my dogs better. Not as good as them, but better than we were. I often like to train with people that raise the bar on my training. I train scent discrimination tracking / trailing with our K-9's. We will take scent from a car's steering wheel or driver's seat after a pursuit and bail. Or from a place that I know a suspect touched after a burglary or robbery. I have been pretty fortunate at tracking suspects through neighborhoods, in high winds and rain or snow to the house or apartment that they were in. My dog has a whole bunch of apprehensions after scent discrimination tracking. I can attribute a lot of this success to the training that I have done with my SAR friends.
This is also interesting and I have seen this study:
I found a blog by one of the authors that talked about a study where something like 15 police k9 teams were sent into a church to look for drugs. Almost all of the teams alerted. All the human officers were told there were likely drugs there. But there were not. The problem of cuing the dog for false alerts is well documented.
Since I train narcotics detection dogs for my dept and other agencies, "false alerts" are an issue. It is certainly a question that gets raised in Court from time to time. I really do not like the term "false alert." The way we train and the way we work, "false alerts" are really not an issue.
The alert is not the trained response of sitting, downing or as I train nose to source and focused stare at the source of odor for a passive dog. Nor is the "alert" the scratching or biting at source for an active or aggressive alerting dog. In either case, passive or aggressive, the alert is the behavioral changes and the physiological changes the dog exhibits before alerting. With every dog doing detection work, the handler must look for the breathing change, the "head snap", the ear and tail set changes, body languages changes and the increased respiration that shows the dog is in odor. These are the things that the alert is based on, not the sit, stare or scratch. A "sit" or "scratch" with out the breathing change, behavioral changes and body language changes is not an alert.
A handler can "cue" a dog to sit or scratch, but you can not "cue" a behavioral change, a "head snap" or an increase in respiration from about 30 bpm to 100 - 150 bpm. This is what handlers need to be trained to realize, look for and base alerts on. When this is done there are zero "false alerts."
I set up training scenarios regularly like the one described, pressure the handler to call an alert, when he/she doesn't call an alert in a blank room for example and goes to walk out I say "are you sure that you don't want to check the room again?" "Maybe, you should check the desk?" Out handlers know to trust the dog, the dog has the nose and the super senses, and not fall victim to the added pressure and stress we put on. Running aids blind is the best way, that way the handler has to rely on the dog. Same as running blind tracks are the only way to really learn to read a tracking dog.
“Beware of false knowledge; it is more dangerous than ignorance”. George Bernard Shaw