Detection Dogs: Are They Reliable? (Part 1)
Dogs can detect incredibly faint odors. But what if their human handlers give them the wrong cues? The Dog Trainer explores dangerous territory.
Page 2 of 2
How the Experiment Worked
To influence the handlers’ expectations, they were told that paper markers would indicate the presence of a target scent. Also, the experimenter had a metal box with 12 triple-bagged half-ounce samples of pot – I’ll just give you a moment with that thought – and a canvas bag with 12 triple-bagged half-ounce samples of gunpowder. These containers were left by the door. The experimenter never handled the samples, and the containers were never opened inside the church.
Inside the testing area, there was no target scent at all, though in some spots there was a decoy scent – that of Slim Jims or of a brand-new tennis ball. So there were four different situations: no decoy scent and no paper marker; paper marker but no decoy scent; decoy scent but no paper marker; and decoy scent with paper marker. Remember that despite what the handlers were told, no place in the testing area had the scent of drugs or of explosives. This meant that any alert was a false alert.
225 False Alerts!
The space allowed for 4 search areas, and each team searched each area twice, for a total of 164 searches. And I hope you’re sitting down for this next bit: Of those 164 searches, 21 – or 15% – were clean. The handlers in those searches didn’t call an alert. The other 123 searches, 85%, resulted in at least one alert. And all told, there were 225 alerts. I’ll remind you again that there was no target scent in the search area.
Every single one of those alerts was wrong!
One important point is that although handlers called alerts both when the paper marker meant to mislead them was present, and when the decoy scent was present, the paper markers elicited more alerts – that is, the human’s expectation that a target scent was present had a bigger effect than any interest the dog might happen to show in a decoy scent.
Why So Many False Alerts?
Everybody involved in this experiment seems to have been flabbergasted. I certainly was, and I bet you are too. Why so many false alerts?
It could have been the Clever Hans effect – the handlers were unconsciously cuing the dogs to alert in places where they believed scent was present. Or the handlers could have mistakenly perceived the dog as alerting, because the handler believed a target scent was present. Or the handlers’ belief that scent was present could have led them to say their dog was alerting even if he or she wasn’t. And three handlers, winning no ethics prizes, frankly admitted that they cued their dogs to alert in the locations with paper markers. Unfortunately, the searches weren’t video recorded. That doesn’t weaken the results of the study, but of course it does mean that we can’t analyze the handlers’ and dogs’ behavior closely.
I imagine that’s enough to think about for now. Next week, I’ll explore objections to Professor Lit’s study as well as some training fixes that might prevent many false alerts. I’ll also step off my Dog Trainer patch to point you to some discussions of the social and legal implications of mistakes by scent-detection teams.
As always, you can write to me at email@example.com. I get so many questions that I can’t respond individually, but check out past episodes – I might already have answered yours, especially if it’s about housetraining, humping, or aggression. And please visit me on Facebook, where I’m The Dog Trainer.
Betterment LLC is an SEC Registered Investment Advisor. Brokerage services are offered by Betterment Securities, an SEC registered broker-dealer and member FINRA/SIPC. Investments are not FDIC Insured. No Bank Guarantee. May Lose Value. Investing in securities involves risks, and there is always the potential of losing money when you invest in securities.