Drugs or No Drugs, K-9s May Alert to Your Car Nearly Every Time

[VMI77]

or the handler is causing the dog to alert on purpose even if the dog doesn't hit for real(yea cops would NEVER lie their buts off to get what they want)

This is what I believe to be the case.

One of my professors brought in his dog at the end of the semester. He had all of us exit the classroom and he hid a training device for the dog to search. Once we were all back in, he told the dog, "Go find dope." and the dog found the device rather quickly.

With all the training provided for the dogs, I just don't believe a 40% inaccuracy rate.

So the evidence before the court was wrong or what? The 40% inaccuracy rate was determined by evidence presented in a criminal trial in Federal Court.

The evidence is correct, however, it puts all the blame on the dog. I believe the handler needs to be investigated as well.

Since a dog doesn't have free will I'd say all the blame falls on the humans responsible for training, certifying, and handling him.

[VMI77]

or the handler is causing the dog to alert on purpose even if the dog doesn't hit for real(yea cops would NEVER lie their buts off to get what they want)

This is what I believe to be the case.

One of my professors brought in his dog at the end of the semester. He had all of us exit the classroom and he hid a training device for the dog to search. Once we were all back in, he told the dog, "Go find dope." and the dog found the device rather quickly.

With all the training provided for the dogs, I just don't believe a 40% inaccuracy rate.

So the evidence before the court was wrong or what? The 40% inaccuracy rate was determined by evidence presented in a criminal trial in Federal Court.

I do not know exactly how the evidence was presented or what that evidence entailed. With that said, it is possible that the evidence wasn't necessarily wrong, but may have been presented poorly. It is important to know whether that 40% inaccuracy claim comes from a controlled test, in which the K9s failed, or if the number comes from real world events in which 40% of the examined cases yielded no drugs. If it is the former, then it can be reasonably concluded that K9s should not be used for such purposes. However, if it is the later, then there are too many uncontrolled variables, such as the K9 picking up cues from the handler, to conclude that K9s themselves are inaccurate.

Seems like it should be the opposite to me.....if it works in the lab but not in the field it doesn't work. Dogs aren't searching people for drugs in a lab, they're doing it in the real world and the consequences to those being searched are real consequences. As the court said, a 40% inaccuracy rate is near random and close to being achieved with a coin toss. A search with a 40% false positive rate is not a probable cause search...it's closer to a just go ahead and search everybody search. But the flip side of this is that he signaled for drugs 93% of the time. Essentially that means you've got less than a 1 in 10 chance of not being searched. That defeats the intent of requiring probable cause.

[mojo84]

What restitution is offered to the person that encounters damage from a search triggered by a false positive hit from a dog and it's handler? Until there is accoutability, nothing will change.

[Beiruty]

What restitution is offered to the person that encounters damage from a search triggered by a false positive hit from a dog and it's handler? Until there is accoutability, nothing will change.

Sue the city.

[mojo84]

What restitution is offered to the person that encounters damage from a search triggered by a false positive hit from a dog and it's handler? Until there is accoutability, nothing will change.

Sue the city.

I understand that's a possibility. However, it is a costly, time and money, proposition. It's not always the city that does the search. Wish it was that simple.

[uthornsfan]

I learned from our local Police Department, that K-9 (sniffing dogs has to go through a certification process).
In the demonstration at Police Citizens Academy, The sniffing K-9 dogs were amazing and accurate. Also as for defense dog duties they were controllable and obedient to commands (issued in German). One thing, I noted when dog attacked the suspect, getting him off was a bit problematic ( the handler has to yell many times to get the dog attention, understandable as the dog was doing a serious business )

If you have ever done anything that requires serious focus (playing sports, hunting, shooting etc.) you will know that some of your senses shutdown. A dog is very similar. If you watch Cesear Milan you will see that alot of times he touches the dog to distract them. That would be highly dangerous when a dog is bring down a suspect.

How many people have been bitten trying to break up dog fights because they reach in and grab the dog :)

You either need a stick or stitches!

[WTR]

i've beeh through Border Patrol check points hundreds of times and have been checked by their dogs. I have never had a hit. The BP asked if they could put a box of weed into my horse trailer once as a training aid. The dog did hit on the trailer.

[Beiruty]

I learned from our local Police Department, that K-9 (sniffing dogs has to go through a certification process).
In the demonstration at Police Citizens Academy, The sniffing K-9 dogs were amazing and accurate. Also as for defense dog duties they were controllable and obedient to commands (issued in German). One thing, I noted when dog attacked the suspect, getting him off was a bit problematic ( the handler has to yell many times to get the dog attention, understandable as the dog was doing a serious business )

If you have ever done anything that requires serious focus (playing sports, hunting, shooting etc.) you will know that some of your senses shutdown. A dog is very similar. If you watch Cesear Milan you will see that alot of times he touches the dog to distract them. That would be highly dangerous when a dog is bring down a suspect.

How many people have been bitten trying to break up dog fights because they reach in and grab the dog :)

You either need a stick or stitches!

I understand, as I had once the narrow field of view and loss of hearing. No wonder why the dog would not hear his handler.

[jkurtz]

or the handler is causing the dog to alert on purpose even if the dog doesn't hit for real(yea cops would NEVER lie their buts off to get what they want)

This is what I believe to be the case.

One of my professors brought in his dog at the end of the semester. He had all of us exit the classroom and he hid a training device for the dog to search. Once we were all back in, he told the dog, "Go find dope." and the dog found the device rather quickly.

With all the training provided for the dogs, I just don't believe a 40% inaccuracy rate.

So the evidence before the court was wrong or what? The 40% inaccuracy rate was determined by evidence presented in a criminal trial in Federal Court.

I do not know exactly how the evidence was presented or what that evidence entailed. With that said, it is possible that the evidence wasn't necessarily wrong, but may have been presented poorly. It is important to know whether that 40% inaccuracy claim comes from a controlled test, in which the K9s failed, or if the number comes from real world events in which 40% of the examined cases yielded no drugs. If it is the former, then it can be reasonably concluded that K9s should not be used for such purposes. However, if it is the later, then there are too many uncontrolled variables, such as the K9 picking up cues from the handler, to conclude that K9s themselves are inaccurate.

Seems like it should be the opposite to me.....if it works in the lab but not in the field it doesn't work.

Not necessarily.

The purpose of lab testing is to find out what conditions are required for success, as well as isolate points of failure. In other words, ABC works under XYZ conditions.

In the case of drug sniffing K9s, it is import to know whether or not K9s are capable of getting accurate hits in a lab. If the K9s are inaccurate in the lab, where they are the only possible point of failure, then the practice of using K9s to detect drugs should be done away with. On the other hand, if they are accurate in the lab, but not in practice where other failure points exist, it doesn't mean that the K9s ability to detect drugs does not work. Rather, it demonstrates that further testing is required in order to isolate the point of failure.

That is why it is important to know where the 40% figure comes from. Did it come from a lab where the K9 was the only point of failure, thus concluding that K9s are incapable of accurate drug detection? Or did it come from an examination of real scenarios where the K9 may have been influenced to give an alert in the absence of drugs.

Finding the point of failure is critical, but not just for drugs. K9s are also often used for explosive detection in the US and abroad. If K9s are not capable of providing accurate hits based on scent, then we need to find better methods of improving security. However, if K9s are capable of providing accurate hits, then we should further examine the role handlers play in detection, or lack there of.

[VMI77]

I learned from our local Police Department, that K-9 (sniffing dogs has to go through a certification process).
In the demonstration at Police Citizens Academy, The sniffing K-9 dogs were amazing and accurate. Also as for defense dog duties they were controllable and obedient to commands (issued in German). One thing, I noted when dog attacked the suspect, getting him off was a bit problematic ( the handler has to yell many times to get the dog attention, understandable as the dog was doing a serious business )

If you have ever done anything that requires serious focus (playing sports, hunting, shooting etc.) you will know that some of your senses shutdown. A dog is very similar. If you watch Cesear Milan you will see that alot of times he touches the dog to distract them. That would be highly dangerous when a dog is bring down a suspect.

How many people have been bitten trying to break up dog fights because they reach in and grab the dog :)

You either need a stick or stitches!

One of our dogs was attacked by bigger dog. When it looked like our dog was in big trouble my son intervened and had to reach in to break up the fight and save him. He got bitten. Wasn't too bad, but did require some stitches.

I've tried some of the Cesear Milan stuff with our dogs but most of the time I can't manage to pull it off like he does. We have a German Shepherd mix that is absolutely relentless once she gets focused on doing something.

[Pariah3j]

i've beeh through Border Patrol check points hundreds of times and have been checked by their dogs. I have never had a hit. The BP asked if they could put a box of weed into my horse trailer once as a training aid. The dog did hit on the trailer.

I could believe this, however the problem is as has been stated above, is ultimately with the handler. So if you have a handler giving bad/false queues whether intentional or not, it can skew the results.

[VMI77]

or the handler is causing the dog to alert on purpose even if the dog doesn't hit for real(yea cops would NEVER lie their buts off to get what they want)

This is what I believe to be the case.

One of my professors brought in his dog at the end of the semester. He had all of us exit the classroom and he hid a training device for the dog to search. Once we were all back in, he told the dog, "Go find dope." and the dog found the device rather quickly.

With all the training provided for the dogs, I just don't believe a 40% inaccuracy rate.

So the evidence before the court was wrong or what? The 40% inaccuracy rate was determined by evidence presented in a criminal trial in Federal Court.

I do not know exactly how the evidence was presented or what that evidence entailed. With that said, it is possible that the evidence wasn't necessarily wrong, but may have been presented poorly. It is important to know whether that 40% inaccuracy claim comes from a controlled test, in which the K9s failed, or if the number comes from real world events in which 40% of the examined cases yielded no drugs. If it is the former, then it can be reasonably concluded that K9s should not be used for such purposes. However, if it is the later, then there are too many uncontrolled variables, such as the K9 picking up cues from the handler, to conclude that K9s themselves are inaccurate.

Seems like it should be the opposite to me.....if it works in the lab but not in the field it doesn't work.

Not necessarily.

The purpose of lab testing is to find out what conditions are required for success, as well as isolate points of failure. In other words, ABC works under XYZ conditions.

In the case of drug sniffing K9s, it is import to know whether or not K9s are capable of getting accurate hits in a lab. If the K9s are inaccurate in the lab, where they are the only possible point of failure, then the practice of using K9s to detect drugs should be done away with. On the other hand, if they are accurate in the lab, but not in practice where other failure points exist, it doesn't mean that the K9s ability to detect drugs does not work. Rather, it demonstrates that further testing is required in order to isolate the point of failure.

That is why it is important to know where the 40% figure comes from. Did it come from a lab where the K9 was the only point of failure, thus concluding that K9s are incapable of accurate drug detection? Or did it come from an examination of real scenarios where the K9 may have been influenced to give an alert in the absence of drugs.

Finding the point of failure is critical, but not just for drugs. K9s are also often used for explosive detection in the US and abroad. If K9s are not capable of providing accurate hits based on scent, then we need to find better methods of improving security. However, if K9s are capable of providing accurate hits, then we should further examine the role handlers play in detection, or lack there of.

I agree it is necessary to prove them under controlled conditions first. However, the most relevant number here isn't the 40% false positive rate. And yes, the rate is the product of the dog and handler working together. The most relevant number is that the dog signaled for drugs 93% of the time. Maybe that wouldn't have happened with a different handler...but as far as this particular dog and handler go it effectively means that everyone got searched, so the searches weren't really based on probable cause.

I suspect the handler was the bigger influence. In the case of explosive detection dogs, the handler has no incentive to trigger false positives, and in fact, some potentially strong disincentives. When it comes to drug detection dogs there are large incentives to trigger false positives because there is a financial benefit if drugs are found and virtually no downside if they aren't. In fact, the incentives are practically reversed....there is no benefit to false positives in explosive detection and no benefits for true negatives in drug detection.

[jkurtz]

or the handler is causing the dog to alert on purpose even if the dog doesn't hit for real(yea cops would NEVER lie their buts off to get what they want)

This is what I believe to be the case.

One of my professors brought in his dog at the end of the semester. He had all of us exit the classroom and he hid a training device for the dog to search. Once we were all back in, he told the dog, "Go find dope." and the dog found the device rather quickly.

With all the training provided for the dogs, I just don't believe a 40% inaccuracy rate.

So the evidence before the court was wrong or what? The 40% inaccuracy rate was determined by evidence presented in a criminal trial in Federal Court.

I do not know exactly how the evidence was presented or what that evidence entailed. With that said, it is possible that the evidence wasn't necessarily wrong, but may have been presented poorly. It is important to know whether that 40% inaccuracy claim comes from a controlled test, in which the K9s failed, or if the number comes from real world events in which 40% of the examined cases yielded no drugs. If it is the former, then it can be reasonably concluded that K9s should not be used for such purposes. However, if it is the later, then there are too many uncontrolled variables, such as the K9 picking up cues from the handler, to conclude that K9s themselves are inaccurate.

Seems like it should be the opposite to me.....if it works in the lab but not in the field it doesn't work.

Not necessarily.

The purpose of lab testing is to find out what conditions are required for success, as well as isolate points of failure. In other words, ABC works under XYZ conditions.

In the case of drug sniffing K9s, it is import to know whether or not K9s are capable of getting accurate hits in a lab. If the K9s are inaccurate in the lab, where they are the only possible point of failure, then the practice of using K9s to detect drugs should be done away with. On the other hand, if they are accurate in the lab, but not in practice where other failure points exist, it doesn't mean that the K9s ability to detect drugs does not work. Rather, it demonstrates that further testing is required in order to isolate the point of failure.

That is why it is important to know where the 40% figure comes from. Did it come from a lab where the K9 was the only point of failure, thus concluding that K9s are incapable of accurate drug detection? Or did it come from an examination of real scenarios where the K9 may have been influenced to give an alert in the absence of drugs.

Finding the point of failure is critical, but not just for drugs. K9s are also often used for explosive detection in the US and abroad. If K9s are not capable of providing accurate hits based on scent, then we need to find better methods of improving security. However, if K9s are capable of providing accurate hits, then we should further examine the role handlers play in detection, or lack there of.

I agree it is necessary to prove them under controlled conditions first. However, the most relevant number here isn't the 40% false positive rate. And yes, the rate is the product of the dog and handler working together. The most relevant number is that the dog signaled for drugs 93% of the time. Maybe that wouldn't have happened with a different handler...but as far as this particular dog and handler go it effectively means that everyone got searched, so the searches weren't really based on probable cause.

I suspect the handler was the bigger influence. In the case of explosive detection dogs, the handler has no incentive to trigger false positives, and in fact, some potentially strong disincentives. When it comes to drug detection dogs there are large incentives to trigger false positives because there is a financial benefit if drugs are found and virtually no downside if they aren't. In fact, the incentives are practically reversed....there is no benefit to false positives in explosive detection and no benefits for true negatives in drug detection.

I would argue that the 40% failure rate is far more important than the 93% of the times the K9 signaled. If the K9 and handler are working together properly, then the frequency at which the K9 signals is just a product of how many people have drugs. Basically, the number of times something happens and fails is far more important than the number of times it happened.

Although there is no incentive to giving a false positive for explosive detection, it was pointed out by others that K9s do at times pick up on unintentional behaviors made by the handler. For example, if a dog is being used to find IEDs in an area and the handler already has a strong belief that IEDs are present in that given area, will the dog alert to an item that the handler is suspicious of, regardless of explosive content?

[JALLEN]

One problem with these situations is that you never get to cross examine the dog.

Having had the disquieting experience of police officers perjure themselves, I remain skeptical of these adventures.

O/T. My siblings and I recall with amusement the time my grandmother went out to break up a fight between her dog and an intruder dog. The intruder dog bit her, and died. The ladies in her church circle meetings had more fun with that, and tormented my grandmother over it, for the rest of her life, I think.

[VMI77]

One problem with these situations is that you never get to cross examine the dog.

Having had the disquieting experience of police officers perjure themselves, I remain skeptical of these adventures.

O/T. My siblings and I recall with amusement the time my grandmother went out to break up a fight between her dog and an intruder dog. The intruder dog bit her, and died. The ladies in her church circle meetings had more fun with that, and tormented my grandmother over it, for the rest of her life, I think.

On a smaller scale, when I was about 10 years old I had a hamster. I reached into his cage one day to pick him up and he bit me. He died later than night. It was a long time before anyone in the family forgot that incident.