Phd Study

[corvus]

How variable is 'amicability', really? There's no point in going into so much detail that you can only categorise one or two dogs under the heading "amicable". As far as I'm concerned, if the dog sees a complete stranger and pins his ears back, wags his tail low and approaches the stranger, he's amicable. How much detail you go into will depend on how many categories under the heading "amicable" you want, and that is dependent on how many would be most useful to the average dog owner. If my dog-loving partner is anything to go by, when they say they want a friendly dog, they mean they just want a dog that will do the above. You wouldn't categorise a dog entirely on that one characteristic, but you could put a tick in the box that says " very amicable" or "moderately amicable" and go from there. With a set of scientifically produced protocols on different traits, you could go a long way to matching an adult dog to the right family.

I don't see that it matters whether the dog is amicable inherently or by the way it was raised. It seems to be beside the point to me. It's a hard thing to separate, so why bother trying unless it's one of the aims of your project? Isn't the whole point of looking at adult dogs to dodge that difficult issue? Besides which, a protocol that is dependent on a condition that doesn't really exist (a dog in isolation) is useless. The protocol should apply regardless of the origins of the behaviour to be useful.

[mlc]

How variable is 'amicability', really? There's no point in going into so much detail that you can only categorise one or two dogs under the heading "amicable".... You wouldn't categorise a dog entirely on that one characteristic, but you could put a tick in the box that says " very amicable" or "moderately amicable" and go from there. With a set of scientifically produced protocols on different traits, you could go a long way to matching an adult dog to the right family.

I don't see that it matters whether the dog is amicable inherently or by the way it was raised. It seems to be beside the point to me. It's a hard thing to separate, so why bother trying unless it's one of the aims of your project? Isn't the whole point of looking at adult dogs to dodge that difficult issue? Besides which, a protocol that is dependent on a condition that doesn't really exist (a dog in isolation) is useless. The protocol should apply regardless of the origins of the behaviour to be useful.

Tammie's research project is about whether an assessment can accurately and reliably (in the scientific sense of those words) assess amicability. This means will an assessment prove accurate (remember Tammie is getting owners to complete 3 questionnaires to report on the way they perceive their dog's behaviour) when repeated at different times (test-retest reliability) and by different people (inter-rater reliability).

The research is actually about the assessment protocol methodology, not about 'labelling' dogs. It also about how experienced dog people DO ascertain a dog is amicable. Is it as simple as tail carriage, ear position and approach (that can be taught through education) or is there more to it? Do you also note lips/teeth/cheeks/pace/hackles/etc.? Is it context specific (known environment compared to novel environment)? Does it differ according to the 'stranger' (eg. children vs men vs women vs someone wearing a hat, etc.)?

Although it would be lovely to have a comprehensive assessment for all elements of canine personality that is short, able to be repeated, totally reliable and accurate, so far the science suggest this is still to come. Logic suggests scientist start by trying to define just one of those known key canine personality traits. Australian's who participated in the scoping survey reported amicability is most important to them - so it makes sense to attempt that assessment protocol first.

This project is focussing on adult dogs, because puppy tests have proved (at this stage) unreliable predictors of adult behaviour. Although their reliability improves with age of the pup, the most stable way to develop an assessment protocol is to use adult dogs.

I'm not sure where the 'dog in isolation' bit has come from. If you refer to the part of the protocol where the owner leaves the dog unattended for a short period, it's very interesting to see how the dog's behaviour changes in the absence of the owner and in a way the owner may not have any idea about. This is easily replicated in 'real life' scenarios and certainly pertinent and a condition that does exist in reality.

Again - hope this helps clarify!

[Jed]

lilli

I'm sure you would be aware, that some of the functional traits selected for back in ye olde

required behavioural charactertics in order for the dog to complete its task properly?

And those behavioural characteristics are what makes a breed satisfactory to a particular owner, without being used for "traditional purposes". I intensely dislike dogs which show fear of fireworks and storms. That makes the Cocker Spaniel an ideal choice -they were bred to stand for the guns, they were bred with long ear leathers to muffle the sound of guns beside them. Genetically, they are programmed not to react to loud bangs - guns, storms, fireworks.

The termperament of a dog is defined as the sum of all its inherited and acquired physical and psychologicl traits, characeterists and abilities, wich determine shape and regulate the dogs responses to its environment (Seiferle, 1972)

The fact that environment, a variable difficult to control and standardize, is part of its definintion makes the experimental investigation of temperament differences between breeds extremely difficult, since such experiments must bebased on the objective temperament evaluation. (Stur, 2000)

Comparison of DNA sequences among members of different dog breeds revealed that there are high levels of genetic variability within breeds. The founding stock of our modern dog breeds was likely drawn from a large and genetically diverse pool of dogs (Dennis-Bryan and Clutton-Brock 1988). Many our our modern dog breeds were created by cross-breeding. Second dog breeds were actually not as ightly inbred during development of our modern breeds as it was assume. Although the goal of high phenotypic uniformity within a breed led to closed gene pools at some point during the development of modern breeds, dog breeders occasionally outcrossed their purebred dogs to avoid negative effects of inbreeding on health and behaviour (Ubbin et al 1992)

(Dr Cornelia Wagner DVM, MS)

[corvus]

The research is actually about the assessment protocol methodology, not about 'labelling' dogs. It also about how experienced dog people DO ascertain a dog is amicable. Is it as simple as tail carriage, ear position and approach (that can be taught through education) or is there more to it? Do you also note lips/teeth/cheeks/pace/hackles/etc.? Is it context specific (known environment compared to novel environment)? Does it differ according to the 'stranger' (eg. children vs men vs women vs someone wearing a hat, etc.)?

Sorry, I wasn't trying to criticise. I was trying to make the point that if you're going to look at a personality trait like amicability, you inevitably have to either create categories or scores of amicability, and you may choose to gloss over some of the details because they aren't important. For example, like you say, maybe tail carriage is important but maybe pace is not (just as an example), so you decide to ignore pace even though different dogs will approach at different paces. Someone made the point that lots of different people see amicability as different things, and I was trying to say that I think it's the opposite. I think it only need be a very broad definition to suit what the average dog owner seeks.

Australian's who participated in the scoping survey reported amicability is most important to them - so it makes sense to attempt that assessment protocol first.

Makes perfect sense to me. I have an undeniably amicable dog and complete strangers fall for him all the time. I love watching how easy it is. He need only sidle up to someone and look into their face with his ears pinned back and they start talking to him and telling him what a lovely dog he is. I think there's a little more to it than just his willingness to engage with strangers, as he does so in a very calm and gentle manner without jumping up or slobbering and I think that's more universally appealing, but not many people ignore him or brush him off when he appears at their side and keeps pace with them as though he is their dog not mine.

This project is focussing on adult dogs, because puppy tests have proved (at this stage) unreliable predictors of adult behaviour. Although their reliability improves with age of the pup, the most stable way to develop an assessment protocol is to use adult dogs.

I would say that's where the complications of nature versus nurture come in? You have to either deal with that or dodge it. I think focusing on adult dogs is a perfectly legitimate way to dodge it. Especially considering dealing with it adds unnecessary complexity to the whole thing.

I'm not sure where the 'dog in isolation' bit has come from. If you refer to the part of the protocol where the owner leaves the dog unattended for a short period, it's very interesting to see how the dog's behaviour changes in the absence of the owner and in a way the owner may not have any idea about. This is easily replicated in 'real life' scenarios and certainly pertinent and a condition that does exist in reality.

Sorry, my bad. I was referring to the difficulties of dealing with nature versus nurture. I can't remember who, but some people seemed to be making the point that by looking at adult dogs Tammie was ignoring the fact that how a dog is raised has even more impact on their temperament than their genetics (arguably). But isolating nature from nurture is notoriously difficult, and would require raising a dog in isolation (which I believe has been done, and they proved that dogs find humans terrifying if they never saw one as a youngster), and even then that doesn't really prove anything. I think to really make that distinction between nature and nurture without genetics you would need to test the puppy at 6 weeks, and probably then again at 8 weeks, and then again at several times until maturity. It would never be clearcut, because behaviour isn't and that's what's so interesting about it, but it might give you a good indication of to what extent the behaviour changes from puppyhood to adulthood.

Sorry, I think I'm rambling a bit. Anyway, my point was it doesn't matter whether the dog is amicable because of genetics or because it has learnt to be because there are no situations in real life where the dog is behaving entirely as a result of genetics, except for perhaps puppy mill dogs. I think I'm not communicating this point very well. What I'm trying to say is it's a moot point and Tammie is therefore right to not attempt to address it. You can only work with what you've got. I think in many cases where this hypothetical protocol would be applied the history of the adult dog would be unknown, so all you can do is assess what's in front of you at the time.

[WoofnHoof]

I think one of the things that gets me confused about this particular field of research is that all of these follow on studies are based on the results of a survey of what Australian's 'say' they want in a dog, whereas a far more accurate assessment of what people really want in a dog would be to measure the traits present in those dogs which have gone the distance, as in lived the full term of their natural lives with their owners as opposed to dogs which were purchased and then gotten rid of.

Hi again,

hoping to offer some clarification...

'all of these follow on studies' = Tammie's one current research project to develop an assessment protocol. The survey cited in my earlier post was conducted to set the direction for her PhD research.

Keep in mind the general Australian public might know what they want in a dog - I wouldn't want to write them off without consideration for their point of view, would you?

And again - I feel I need to emphasise that just because it may be different to someone's individual point of view, doesn't make it wrong or not valid.

If Tammie doesn't use the results of what almost 1,000 people reported they wanted in a companion pet, then what do you propose she use to guide her? The survey did ask people about their current dog, their level of satisfaction with that dog and took into consideration if the reported 'ideal' dog matched with the 'profile' of their current dog, so it did capture information about successful matches, but also revealed a lot about unsuccessful ones. As do the shelter figures.

Lol I would totally write the Australian public off they are notoriously awful at knowing what they actually want, just look at our pollies

I do see your (and Tammie's) point but I would rely far more heavily on the data from the successful (and unsuccessful) matches than anything else, the proof being in the pudding and all that. Similarly people who have completed breed compatibility tests and ended up with completely different results than the breeds that they currently own and are happy with suggests that either compatibility tests are faulty (possible) or that people's perceptions of what they really want in a dog are faulty (also possible).

[mlc]

But isolating nature from nurture is notoriously difficult, and would require raising a dog in isolation (which I believe has been done, and they proved that dogs find humans terrifying if they never saw one as a youngster), and even then that doesn't really prove anything. I think to really make that distinction between nature and nurture without genetics you would need to test the puppy at 6 weeks, and probably then again at 8 weeks, and then again at several times until maturity. It would never be clearcut, because behaviour isn't and that's what's so interesting about it, but it might give you a good indication of to what extent the behaviour changes from puppyhood to adulthood.

You might be interested to know that Tammie proposes to test a group of pups and then retest them 12 months later. Should be fun! She just needs some volunteers. If you are interested in participating, you can contact her at: [email protected]

Another member of the Anthrozoology Research Group (Tiffani Howell) is looking into dog cognition and when pups first start using the social cues that 'engage' humans (like shared gaze in unsolvable task exercises and following human pointing cues, etc.)

Lots of great research under way!

[Danois]

mlc - what is your connection to this? You seem to know a lot about the proposed research and I am just curious.

[Jed]

mlc

following human pointing cues, etc.)

Hasn't that been done? I recently read somewhere that the results of a study identified definitive age for following pointing cues

if so, why is it being replicated?

mlc

To determine this, 877 participants (79.8% female) aged 18–82 years (mean = 34.3, SD = 14.5) were surveyed. A number of behavioural and physical characteristics were identified as important to Australians. These included dogs being medium sized, short haired

Eurodog.

Interesting that the results should show that, yet the public snaps up every Cavallier, Cocker, Poodle, Pomeranian, Bichon and other hairy dogs, as well as hairy oodles in hundreds of thousands. Very few puppy farms breed medium short haired dogs because there is not as much market for them as long haired dogs.

It is well known in pet shops that fluffy puppys (any sort) sell better than anyrhing else, and at a higher profit margin.

Edited January 4, 2010 by Jed

[Erny]

Your power of memory never ceases to amaze me, Jed.

Dog age for following pointing cues

[Jed]

Thanks Erny. Trouble is, only some of the cells work - I know I've read something, but I can never remember where!! And I have extreme difficulty finding it again!!

Appreciated

[mlc]

mlc - what is your connection to this? You seem to know a lot about the proposed research and I am just curious.

Hi Danois,

I'm a member of the Anthrozoology Research Group and Animal Welfare Science Centre. A research PhD candidate colleague of Tammie's.

[mlc]

Hasn't that been done? I recently read somewhere that the results of a study identified definitive age for following pointing cues

if so, why is it being replicated?

You're right - that article has just recently been published.

However, Tiffani's research is already under way and doesn't just use that one assessment task, she's looking at other elements as I outlined. Repeating a test that already has results in conjunction with other new assessments is generally sound science. One study with one methodology from one place with one group of dogs does not necessarily a solid conclusion make! It's gathering up pieces of the puzzle that allow us to put them together and see the bigger picture. Sometimes different groups find different results that reflect changes in their methodology. Repeated findings over time help to suggest the findings are a sound.

The pointing cue task is one of the common cognition tasks in animal cognition research - it's been used with dingoes and wolves and dogs in other research.