13 June 2014

The Kong test - dog

Image: The Kong test to determine the handedness of a dog—A: Right paw used, B: both paws used, 
C: left paw used. Fifty left or right paw uses were recorded to determine paw preference. 
Image courtesy: Dr Lisa Williams.

The commitment and loyalty of a guide dog is second to none—these animals show unconditional love for their owner, but are also carefully trained to care for and enhance the mobility of their owner.

What many may not realise is the amount of time, effort and resources that are channelled into the training of a canine to become a guide dog…and not all dogs make the cut.

New research in this area could make the task of assessing appropriate dogs easier and cheaper.

Dr Lisa Williams (nèe Tomkins) is an animal scientist who conducted her postdoctoral studies under the supervision of ARC grant recipient and Australian Museum Eureka Prize (AMEP) Winner, Professor Paul McGreevy. Currently employed by Guide Dogs NSW/ACT, her PhD project focussed on objective measures that can be used to determine the suitability  of dogs for guiding work.

“In my research I looked at 11 different tests which could be used to objectively measure a dog’s temperament and suitability for guide  dog work,” said Dr Williams.

“Of these tests, those which looked at lateralisation (the left or right handedness of dogs) and hair whorls produced some of the most interesting results.”

Dr Williams assessed 114 dogs that entered a guide dog training program in Sydney. On these dogs, she conducted tests to determine whether they were left or right handed (or pawed). One of these tests involved the use of a common ‘Kong’ toy filled with dog treats. By watching the dogs as they played with the toy and extracted the food from it, on 50 occasions, it could be determined with which paw they showed a preference. Surprisingly, about half of all dogs turned out to be ambidextrous (47%), while 29% of the dogs were left and 24% right handed. 

The next test to determine the handedness of dogs was the ‘first-stepping’ test. Here a dog on a lead was asked to walk down a small flight of stairs, and the paw they used to take the first step was observed. From this test, which also recorded 50 left or right paw uses, 46% of dogs showed a right preference, 30% showed a left preference, and 24% showed no preference.

When these dogs entered the Guide Dog Training Program conducted at Guide Dogs NSW/ACT Training Centre, and once other factors were accounted for, it was found that the dogs which showed a left handed preference in the Kong test had a significantly lower success rate in guide dog training. Only  38% of these dogs passed training, compared to 64–68% of right handed or ambidextrous dogs.

This finding aligns with previous studies which found that left handed animals were more likely to be fearful, and would withdraw from novel situations. Interestingly, dogs which showed a left preference in the ‘first-stepping’ test were not found to be significantly less likely to pass the guide dog test.

“Previous studies have reported that paw preference changes with task complexity and differing preferences for specific tests may reflect different motivation and action patterns required to perform the task,” Dr Williams said.

“Given the Kong test was based on foraging behaviour, whereas the ‘first-stepping’ test was based on a locomotory behaviour, this may help to explain why a difference in the tests, and hence association with guide dog success, was observed.”

Part of the suite of tests conducted on potential guide dogs was an analysis of hair whorls. These spiral patterns which occur at different positions on the coat of most animals have been discovered in cattle and horses to be indicators of the animals’ temperament. 

Dr Williams’ study confirmed and broadened these findings by showing that dogs with clockwise chest hair whorls were only 29% likely to pass the Guide Dog Training Program compared with a 61% success rate in those that had counter-clockwise hair whorls.

“The association between hair whorls and brain development reflects a common embryonic origin shared by the skin and the nervous system,” said Dr Williams.

“That said, hair whorls are of particular interest as they could potentially be used as an early indicator of guide dog success in young pups prior to entering the training program.

“Training a guide dog costs $30 000 so if these tests could assist selecting pups with a greater chance of becoming a successful guide dog, then significant savings could be made.”

Since completing her PhD, Dr Williams has been working with Guide Dogs NSW/ACT, while Professor Paul McGreevy is continuing his research into animal behaviour and welfare science at the Faculty of Veterinary Science at the University of Sydney.