1.01.2011

The Pit Bull Hoax: The ATTS


There are three kinds of lies: lies, damned lies, and statistics.
Mark Twain

"Pit bulls score higher than ______"
As soon as the topic of pit bulls hits the comment sections, the pit bull apologia goes on the defensive and more often than not, the ATTS (American Temperament Testing Society) is the weapon of choice from their arsenal of myths. Typically the pit bull's ATTS scores will be inflated to unbelievable heights. It is not uncommon to read "pit bulls score higher than ANY other breed." A quick check of the available on demand ATTS statistics will bust that myth straight away. But the pit bull does score higher than many popular, safe, family friendly breeds of dogs. How could the dog responsible for roughly 50% of all fatal attacks, score better than Lassie?

This blog post spells out how and why and touches on the following aspects: The history of the test, the test requirements, the test itself, the evaluation, criticisms of the test, junk science, and examples -- in their own words -- of pit bull owners who acknowledge the test's biases and invalidity.

Introduction
Pit bull advocates have much invested in the ATTS myth. They clutch to these scores and parade them around as though they were the definitive tool that proves that their dogs are not only safe to be in the community but SAFER than most other dog breeds. From the NCRC to Best Friends to Badrap, from Jane Berkey to Diane Jessup to Leslie Haller to Adrien P, they all extol the virtues of the mighty ATTS. They mislead the public and preach the superiority of pit bulls based on this flawed test and its perverted results that are further exaggerated when regurgitated ad nauseam by the pit bull ideologues.

Researching the American Temperament Testing Society (the actual organization) was quite an education in and of itself. Information was difficult to obtain, for example the names of the board of directors and testers were not easy to come by. I finally found the names of the board of directors through the Missouri secretary of state website, but good luck finding the names of testers. Unearthing underground dog fighting information would be an easier task.

Suspicious secret society aside, the problem with the ATTS is complex and any discussion of the problematic statistics should begin with the inherent bias of the test, the testers, the testees and the myths surrounding what exactly this test measures.

schutzhund photography

ATTS History
The temperament test was developed by Alfons Ertelt in 1977. Mr Ertelt was not an animal behaviorist, he worked in the print industry but his passion was dogs and he was involved in schutzhund. (schutzhund is a dog sport that mirrors the training of police dog work and it is dominated by german shepherds) The ATTS test was initially intended to test working dogs for jobs such as police work. The test favors bold dogs, dogs that need to face danger head on without hesitation and fear. Courage was desired and rewarded, timidity was not. The ATTS favors dogs like pit bulls over dogs like collies. It is important to note, the test does not evaluate dogs for "pet" suitability. It comes as no surprise that when you look at the numbers tested by breed, four of the top five breeds (5357 rottweilers, 3038 german shepherds, 1574 dobermans, 968 mixed breeds, and 893 bouviers) excel in schutzhund. Presently, their website states the ATTS was established to "work for the betterment of all breeds of dogs." Mr Ertelt left the ATTS a few years after its inception and in 1983 formed the German Shepherd Dog Schutzhund Club of Los Angeles. In 1990 Carl Herkstroeter, his wife Carolyn, and Harriet Ann Pahlmann and Margaret B Pahlmann incorporated the ATTS into a for profit business in the state of Missouri. One year later, they created the non-profit. Records indicate there has been only a few board changes over the last 20 years.

Test Requirements
Dogs must be at least 18 months of age and on a 6 ft lead. The handler is not allowed to talk to or correct the dog. If the dog fails, it is allowed to retake the test after 5 months, but only one retest is allowed. It is not clear if or how this is monitored and enforced.

Entry level testers are called Apprentice Testers and they must be at least 21 years of age and active in dog sport or employed in the veterinary field for at least the past 5 years. As the testers become more experienced, they can work their way up through the ranks.
Provisional Testers
Temperament Testers
Provisional Chief Testers
Certified Chief Tester
Teaching Chief Tester
Since the testers are required to have a great deal of dog sport experience, the testers are often very involved in their breed clubs, therefore the testers are often evaluating their friends and acquaintances and usually the breed of their choice. In other words, they are testing dogs that they have a vested interest in seeing pass the test.

The Test
The test takes approximately 10 minutes and at least 3 ATTS trained evaluators score the dogs. The test is usually sponsored by breed clubs, most often rottweiler clubs but other large powerful breed clubs like german shepherds, pit bulls and akitas sponsor the test as well. The sponsor can request the test be closed to other breeds or open to all. It is easy to see how a lab or a collie would be at a serious disadvantage at a pit bull sponsored test being evaluated by people who not only don't possess breed specific knowledge of labs or collies but also might harbor resentment towards the other more popular and socially accepted breeds.

The dogs are on a 6 ft leash. The owner/handler is forbidden to speak, give commands or corrections. As the dog progresses through the test, it becomes increasingly more stressful. The dogs' reactions are measured toward a neutral stranger and a friendly stranger. The dog/handler then progress to hidden noises, first the metal bucket with coins then gunshots. From there the dog has an umbrella open suddenly nearby and then walks across plastic sheeting and wire grating. The final phase of the test measures a dog's self protective/aggressive behavior by exposing the dog first to a weirdly dressed non-threatening stranger, then a threatening stranger and finally an aggressive stranger.

If a dog panics and does not recover or if the dog shows strong avoidance or unprovoked aggression, it fails. Re: aggression, breed specific temperament and the prior experience and training of the dogs are taken into account. The website states that aggressive responses during the final phase of the test is okay for a dog with schutzhund training but an untrained husky displaying aggression toward the stranger may fail. It is obvious that these judges possess far too much discretion in these tests. Some pit bull owners report that their pit bulls passed when it launched aggressively at the stranger while other pit bulls have passed when it barely acknowledged the stranger. Passing or failing is completely dependent upon the whims of the testers.

The handlers are advised to come early so they can do a walk through of the test, without their dogs.

The Evaluation
There are a total of 10 subtests and each test is scored by at least 3 testers who rate the dog on a scale of 0-10. All dogs start with a score of 5 on each subtest. Points are added to the score of 5 for the dogs who perform positively and points are deducted for dogs who perform negatively. A score of zero on one test is a failure, even if the dog scored a 10 on each of the other 9 subtests. And oddly enough, a dog would pass if it received a score of 1 on each test. If the testers are not in agreement, the majority rules. Click here for a more in depth explanation of the 0-10 rating scale. Since more often than not, these tests are sponsored by breed clubs, and tested by club officers, there is an inherent bias in the testers to see their breed of choice pass and others fail.

ATTS Tests



Current ATTS scores for pit bull type dogs:
american pit bull 86%
american bulldog 84.8%
american staffordshire 83.9%
bull terrier 90.4%
staffordshire bull terrier 89.6%

Current ATTS scores for a few popular breeds of dogs in America:
cocker spaniel 81.9%
collie 79.7%
beagle 80.3%
chihuahua 71.1%
labrador retriever 92.3%
golden retriever 84.6%
german shepherd 84.2%
jack russell 84.1%
mutt 86%
pomeranian 75.8%
pug 90.9%
standard poodle 86%

Of the 30,000+ dogs tested by the ATTS to date, 82.4% have passed.

*The breed of dog with the overall lowest passing score was the skye terrier at 37.5%.
The most "aggressive" dog in America, yet it didn't make the Clifton report.

Test Criticisms
Anyone with a superficial understanding of scientific method and what constitutes good science, should be able to immediately recognize the inherent problems with this test.

First and foremost is the issue of what exactly does the test measure? The ATTS website claims to measure stability, shyness, aggressiveness, friendliness, protectiveness, self-preservation. In theory, the testers consider the following during the test:
The breed of the dog (hereditary purpose)
The training the dog has received
The dog’s age
The dog’s gender
Whether it has been spayed or neutered
The dog’s physical health (dog in season)
Whether it is a house dog or kennel dog
Yet the pit bull advocates present the stats in such a way as to imply that higher scores equal less aggression and lower scores equal more aggression. According to Herkstroeter, “Just because a certain percentage of dogs in a certain breed fail, this does not necessarily indicate aggression. Dogs fail for other reasons, such as strong avoidance. If you look at our statistics just from a perspective of aggression or non-aggression, they can be very misleading.” Herkstroeter states that 95% of the dogs that fail, do so because they lack confidence to approach the weirdly dressed stranger or walk on the strange surface. The remaining 5% fail because they take longer than 45 seconds to recover from the gunshot or the umbrella. Still pit bull advocates continue to distort the meaning of the test.

Second, as per ATTS website: "Comparing scores with other dogs is not a good idea" and the test "takes into consideration each breed's inherent tendencies". Cocker spaniels are evaluated against a cocker spaniel standard, not against german shepherds (or at least in theory, they are not supposed to), yet pit bull advocates would have you believe that all dogs are evaluated against one another.

The third troubling aspect of this test is not only the lack of a random sample but what appears to be pit bull advocates openly conspiring to groom test candidates and cherry pick only those individual dogs that are likely to pass. This is a conscious decision done for the sole purpose of inflating the scores to improve the pit bull's image.

The fourth major problem is in the inherent bias of the testers. Pit bull owners, breeders and advocates are in the position to pass or fail dogs that are under heavy criticism for what their critics perceive as their innate viciousness. As you will see, the testers have much discretion and a vested interest in the outcomes of the tests and they do not apply the rules fairly or consistently. There is no quality control to ensure that the testers are consistent in how they grade dogs' behavior.

Fifth, the test acknowledges that breed of dog (hereditary of purpose) is factored into the dog's performance and score, yet dogs are not tested in the presence of other dogs. This is especially critical with dogs that were bred to fight.

Sixth, the ATTS apparently does not require papers for purebred dogs. It seems that you can report any breed you like and do not need to provide any registry papers to prove it. One thing I find especially interesting is the flexibility around the issue of purity in pit bulls. When pit bulls attack, they are mixes but when they pass the ATTS, they are purebred, no questions asked. Just last week, Drayton Michaels made a point to say that most pit bulls were in fact not purebred pit bulls. This is another favored defense tactic when pit bull attacks hit the news media. In my experience reading all of the pit bull forums for ATTS information, discussion about the test is non-existent among the game-dog.com demographic. This group of pit bull owners is obsessed with bloodlines and pedigrees, and these dogmen do not even broach the subject of temperament testing.

Seventh, the handlers are familiar with the test, they not only know what to expect, they practice it with their dogs. In the real world, things don't work out that way and much of a dog's reaction can be based on the handler's reaction to the real world "stressful" events.

The test and the testers discriminate against timidity and favor courage and aggression, although they do not openly admit to this. The test does not measure any exact personality trait that can be quantified. In theory, the test measures the dogs' responses to random events, (sometimes heavily practiced random events) and theoretically, the testers are to evaluate dogs within their own breed and not against other breeds. It will become obvious to the reader that the testers are inconsistent with their application of the testing guidelines. In addition, these random events have little to do with the real world events. The organization, the test, the testers and the testees are all motivated by powerful forces, such as BSL and will do anything to achieve their desired outcomes. The test looks good in theory but the humans that apply the theory are self interested and therefore fallible. That fallibility produces unscientific results.

The ATTS temperament test is scientifically invalid and unreliable. The test can not reliably predict how a dog will behave in the real world.

Pit bull apologists love to cite beloved breeds of dogs who score lower than the APBT. But the reality is collie owners are not spending hours online seeking advice from other collie owners on how to pass the test, or spending a year preparing their dogs for this test. Collie mix owners are not quizzing testers about whether or not their dogs can slip in under the radar as purebreds. Collie owners are not prescreening their dogs. Collie owners are not hiring ATTS experienced dog handlers to test their dogs. Collie owners are not cherry picking only those candidates that they think will pass the test. If they did, their breed would score higher than 79.4%. Any breed would do better under these artificial conditions. But collie owners don't view their dogs as a cult religion and the ATTS as a bible. Most people owning normal breeds of dogs view this test as a fun way to spend the morning with their dogs. The pit bull community views this test as a get out jail free card. There is a strong push by pit bull advocates to have dogs who have passed the ATTS eligible for home owners insurance and access to housing that specifically excludes certain breeds. Take for example the military ban on pit bulls, rottweilers and others. In the instance of the Marines, owners of these dogs had 60 calendar days to successfully pass a "nationally recognized temperament test." The City of San Francisco, which has a mandatory pit bull sterilization law, allows pit bulls that pass the ATTS to breed (Section 44.1).


This "truth" design for t-shirts, was created by pit bull talk member mnp13 and is available at cafe press for $25. Please note that mnp13 is a pit bull owner who has witnessed up close and personal how flawed the test is.

In their own words
What pit bull owners say in public is much different from what they say to each other in "private".

The link to Diane Jessup advising fellow thepitbull-place.com members to enter only dogs that will pass the ATTS, was killed after I posted it on craven desires. Please alert me to any dead links and I will replace with pdfs and/or screen shots.
click to view at 100%

Pit bull rescuer, owner and advocate Ellena Thomas of the pacific northwest pit bull rescue explains the ATTS.

Thomas' explanation contradicts this individual's experience. Here pitbulltalk member Tiger describes her ATTS and her violation of the test rules while still passing.

apbtmom76
Some testers are sometimes lenient on some aspects of the test on some dogs. It really depends on the whims and the motives of the tester.

pitbulltalk mnp13
If a schutzhund trained Ambull fails to respond aggressively to the stranger, it passes but if a non trained husky responds it fails. The testers are not evaluating dogs consistently or fairly.

pitbullmamaliz

pbf voodoo on the importance of practicing for the test

Experienced "responsible" pit bull owner, ATTS tester Leslie Haller advises pitbulltalk member hey21jude on the ATTS.
Based on the chief tester's comments, i would guess that this pit bull was tested among a GSD, rott or dobie schutzhund club and is either unaware that dogs without protection training should not be responding aggressively or doesn't care.

new pitbulltalk member tradewind introduces herself to the forum.
It is inappropriate for ATTS testers to be active in BSL.

apbtmom76 describes the appropriate response to the threatening stranger.


apbtmom76 tested one of her dogs and gives advice to another pitbull-chat members.

pbf sarah

gsdbulldog on pre-test jitters.

bulldogbreeds member attitude about temperament tests.
Comforting, isn't it?

2005 Romanwild
They are closing in.

2005 lisa mawson
They set a goal and they achieved it. The pit bulls have surpassed the goldens.

pbf member Maryellen Harwelik (realpitbull.com) advises pblove on passing the umbrella test.

All of this practice feels like like cheating to me. This is like getting all of the answers to the test before hand.

pbf Leslie Haller
Other members chime in and explain that they also will not test their dogs because they know they will fail.

pbf Red

Just something to keep in mind for anyone who wants to enter their dogs...unless the animal in question can be exposed to some stimulation for an extended period of time and don't be bother by it keep him/her away in a vehicle or area that is quiet. Most people "park" their dogs by the testing field, which means that the dogs hear the gun shots, the yelling drunken man and dogs barking while waiting to be tested and that can be a significant amount of time. That is a lot for many dogs to handle, especially if they are not show dogs, dogs who compete in some sports or are used to the usual chaos in similar environments. By the time they enter the test their stress level has increased and moving from one station to the other end up with an animal who is overwhelmed or score low.Most test are held in conjunction with some kind of dog event so that is something to take in consideration.
Red is a dog trainer, experienced pit bull owner and experienced at taking the ATTS. She is also apparently an ATTS handler for hire as well, further insuring that a pit bulldog will pass the test with her confident demeanor and further skewing the test scores.

pbsmiles sarallyn
The pit bull apologia knows EXACTLY what it is doing. Unfortunately, the public is oblivious to what lies beneath the cut and paste propaganda scores.


Beating Lassie's Score
After reviewing this blog post and its many examples, now you know why pit bulls -- responsible for roughly 50% of all fatal attacks -- score better than Lassie. I did not come to this conclusion quickly. I spent about a year researching the ATTS. In addition to reading every ATTS related thread on every pit bull forum on the internet, I also looked into other breed forums. I found nothing on the other breed forums that even remotely resembled the machinations of the pit bull apologia.

Anyone with a high school education should be able to see the fallibilities of the ATTS, yet a few of the gifted and educated, some with advanced degrees still promote the ATTS as a test with "scientific" value.

National Canine Research Council aka the great neocortex
(When caught in a lie, the NCRC deleted this free flyer from their website after I exposed their lie here. Here is the pdf.)