The new kid in dog training town: Do as I Do

Guest post on the realities of Do as I do dog training
By dog behaviour therapist Sedrick van Gronsveld, September 2015

About the author: dog behaviour therapist in Belgium


Sedrick is a reputed behaviour therapist for dogs in Limburg (Belgium). Sedrick’s approach is evidence-based and humane. Sedrick combines years of practical experience with the most respected certifications in the field.

In this article, Sedrick explores the daily realities of Do as I do, a dog training approach that is taking the dog world by storm. Without suggesting we throw the baby out with the bathwater, he invites us to ask critical questions.


The new kid in town: Do as I do

Training a dog can be gratifying and enormously entertaining. But sometimes it takes a long time to teach a complex behaviour sequence; and it can be quite challenging.

Wouldn’t it be wonderful if we could just show the dog what we want him to do? You could teach a complex behaviour chain in a matter of minutes instead of days or weeks. That’s where the new method ‘Do as I do’ comes in, but just like everything else, it has its limitations. In this article we explore how this spectacular new method can help you train your dog.

Up till now, if you wanted to train a dog to, say, open a garbage can, you would ‘shape’ the behaviour: rewarding the dog for looking at the bin, then moving towards the bin, then touching the bin, etc. Until the dog actually opened the bin.

While we can teach a dog a lot of tricks using this method, there is potential in other untapped learning processes. All dogs learn socially, but for a long time we didn’t know quite how much, nor how the process works. So we never consciously used this system to teach dogs. However, everybody who has had an older dog and a new pup will attest to the fact that the new pup seemed to learn the ‘house rules’ a lot faster than the older dog did when he was the only dog in the house.

Recently, scientists began researching this phenomenon; more specifically Adam Miklosi and his research group. They found not only that dogs do indeed learn socially, but that they can learn socially from humans too. Thousands of years of domestication have made sure of that.

We now know that dogs have a left-gaze bias when observing humans – just like we do, by the way. But they do not demonstrate this bias when observing any other animal. The left-gaze bias is present in humans, because humans don’t show emotion symmetrically. The right side of the face shows the most accurate emotion (Mills et. al. 2009) This may indicate that dogs place importance in the expression of human emotion. We also know that dogs do know what that smile of yours means. Dogs can follow the direction you are looking or pointing at (Gàcsi et. al. 2009). Also, when faced with a tough puzzle, they tend to look to a human for help, while wolves would keep trying by themselves (Miklosi et. al. 2003). Dogs are very well adapted to life with humans. All of these skills and more, make them perfect candidates to socially learn from humans.

After learning about this capability, Claudia Fugazza, a researcher and dog trainer, wanted to develop a method to put this knowledge to good use. She came up with ‘Do as I do’. I will not be discussing exactly how you can teach this method. For that I refer you to her book and this article. In the following video, you can see ‘Do as I do’ in action. Instead, I will take you through the advantages and disadvantages of the method.

To understand the pros and cons of ‘Do as I do’, we must understand how dogs imitate. It is not an exact imitation but rather a ‘functional imitation’: not imitating the movements exactly, but accomplishing the same task. So it works best when the behaviour demonstrated has a goal.

The imitation rule is taught by first applying the protocol to three well-known behaviours on verbal cue. In this phase the dog learns the association between the demonstration and the command is made. The dog learns that the demonstration is key in choosing the correct behaviour. If all goes, soon you won’t need to give the command. The demonstration will be enough. That’s when you can add another three (already mastered) behaviours to the mix. Now you have six possible behaviours. Now the dog must really pay attention. And finally, the last phase: the dog must apply the imitation rule to learn a completely new behaviour.

To learn the imitation rule, dog cannot receive ANY other clues of what the behaviour could be, except for your demonstration. That’s why the first three behaviours need to be on verbal cue and verbal cue alone (no pointing, no hand gesture, no hinting, nothing). This means the dog reliably sits even if you ask him whilst blankly staring ahead. The only indication he gets is a verbal command. And that is the crux. Most of us use body movements (consciously or not) to accentuate our request. If you do give let up the tiniest of clue through your body movement, the clue could become the ‘discriminative stimuli’, rather than your demonstration. It then becomes really difficult to teach the dog the imitation rule.

This first rule alone makes ‘Do as I do’ difficult to apply. It is definitely possible, but as a dog behaviour therapist, I wonder how many of my clients could actually reach this level of training? I can’t imagine that many. Dog behaviour therapy tends to be demanding enough without having to learn advanced dog training techniques.

Research has also unveiled that ‘Do as I do’ is particularly effective at teaching ‘object-based tasks’ (Fugazza, C., Miklosi, A., 2014). So its value in assistance dogs could be immeasurable. But with behaviours as sit, or lie down, it wouln’t be significantly more expedient than conventional training (Fugazza 2014).

There are also behaviours you cannot physically teach with ‘Do as I do’. Take the “high five” for instance. You can’t demonstrate a high five upon yourself, as the dog would need to high five himself if he follows the imitation rule strictly. If you high five someone else, the dog would have to high five that other person, not you. Now you see why this behaviour is more easily learned with other methods.

‘Stay’ is also something you cannot demonstrate. Think about it… If you demonstrate ‘doing nothing’, how can the dog know that ‘doing nothing’ is what you are demonstrating? Aside from this problem, the dog actually needs to understand the ‘stay’ in order to sit and stay and observe any ‘Do as I do’ demonstration.

Not pulling on a leash is another thing you couldn’t demonstrate to your dog through ‘Do as I do’. How can you convey NOT doing something in a demonstration? Try this with friends: set up a few toys – one of them red – and sit next to them. Ask your friends to guess what you are doing. Wait and see how long it takes them to figure out that you are “not touching the red toy”.

Don’t get me wrong. I think we are living exciting times: dog behaviour research has taken off, and the implications of all the new “Do as I do” knowledge are awesome. Anyone who uses ‘Do as I do’ will tell you how much fun it is, and how it improves your relationship with your dog: you communicate on a whole new level.

But some journalists and bloggers present the method as being far better than any other out there. First off, teaching the imitation principle of ‘Do as I do’ needs the dog to already have learned behaviours. So you NEED the other methods to be able to teach ‘Do as I do’. Furthermore, if your dog has learned a new behaviour using ‘Do as I do’ you will want to be able to tell your dog to “open the drawer” rather than demonstrating it every time, to ‘put the behaviour on cue’. And there you have it, now you’re using classical conditioning again. Lastly, the imitation is a functional one. If you want to perfect the form of a behaviour, you will need to use shaping to get exactly what you want. I agree ‘Do as I do’ is pretty awesome, but it should be seen as an addition to current dog training methods, NOT as a replacement.

Any thoughts?

Experienced in Do as I do? Using it for practical purposes or just for fun? Think it can replace current methods or reckon it’s a fun complementary method? Share your opinions with us, leave a comment.


  • Mills et. al (2009). Left gaze bias in humans, rhesus monkeys and domestic dogs, Animal cognition
  • Gácsi, M., Győri, B., Virányi, Zs., Kubinyi, E., Range, F., Belényi, B., Miklósi, Á. (2009). Explaining dog wolf differences in utilizing human pointing gestures: Selection for synergistic shifts in the development of some social skills. PLoS ONE 4(8): e6584.
  • Miklósi, Á., Kubinyi E., Topál, J., Gácsi, M., Virányi, Zs., Csányi, V. 2003. A simple reason for a big difference: wolves do not look back at humans but dogs do. Current Biology, 13: 763-766.
  • Fugazza, C. and Miklosi, A. (2014) Should old dog trainers learn new tricks? The efficiency of the Do as I do method and shaping/clicker training method to train dogs, Applied Animal Behaviour Science,
  • Fugazza, C. (2014) Do as I do: using social learning to train dogs, Dogwise, 81 pg.

Contract Sedrick

If you live in Belgian Limburg or nearby and need advice on dog behaviour, you can contact Sedrick on Kwispel therapie.

This entry was posted in Dog training and tagged , , , , , . Bookmark the permalink. Post a comment or leave a trackback: Trackback URL.

2 Trackbacks

Post a Comment

Your email is never published nor shared.

You may use these HTML tags and attributes <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>