I just found this in Applied Animal Behaviour Science journal Vol.88, pp 331-341. It's an interesting article, but I think it misses (or sidesteps) the purpose of the clicker in some common situations. The task was simple cone-touching.
The efficacy of a secondary reinforcer (clicker) during acquisition and extinction of an operant task in horses
J. L. Williams, T. H. Friend, , C. H. Nevill and G. Archer
Department of Animal Science, Texas A&M University, 2471 TAMUS, College Station, TX 77843-2471, USA
Abstract
"Clicker training" is a popularly promoted training method based on operant conditioning with the use of a secondary reinforcer (the clicker). While this method draws from theories of learning and is used widely, there has been little scientific investigation of its efficacy. We used 60 horses, Equus callabus, and assigned each horse to one of six reinforcement protocols. The reinforcement protocols involved combinations of reinforcers administered (primary versus secondary plus primary), schedule of reinforcement (continuous versus variable ratio), and reinforcers applied during extinction (none or secondary). There were no differences (P≥0.11) between horses which received a secondary reinforcer (click) followed by the primary reinforcer (food) and those which received only the primary reinforcer (food) in the number of trials required to train the horses to touch their noses to a plastic cone (operant response). There also were no differences (P≥0.12) between horses which received the secondary reinforcer plus primary reinforcer and those which received only the primary reinforcer in regards to the number of trials to extinction. We conclude that there is no difference in the amount of training required to learn the operant task or in the task’s resistance to extinction between horses that received a secondary reinforcer followed by a primary reinforcer versus horses which received only a primary reinforcer.
I read this too and to my very 'unscientific' mind there were plenty of problems with this study including the learning environment, the number of trials etc.....I think at one point I wasn't even sure if they understood just what exactly they were trying to achieve.
In retrospect this was probably a very unfair thought on my part especially when you take into account that they're trying to tell me something I don't want to hear! Still misses the point though IMHO.
Mantra for the day: Must have an open mind, must have an open mind.......
Hi Francis, I saw that one too. Saw a lot of flaws in it. Basically they were training such a simple, discrete, and unusual behaviour that anyone could have told them they could train it just as well without a click. What I'd like to see, would be a trial comparing click vs. no click when training a canter depart on the correct leg, or a nice bend and collected carriage on the lunge with no side reins or even no lunge rope. The big advantage of the click is that it enables you to mark a behaviour clearly, as it is being offered.
Importantly, they actually write how they clicked immediately AFTER the horse touched the cone, not AS the horse touched the cone, which suggests they didn't quite get the point of the clicker.
Finally, the conclusion of that paper was unjustified, although what I read was a draft so I don't know if they left it in the published version? Whereas the research compared using a food reward with a clicker to using a food reward without a clicker, the conclusion went on to suggest that clicker training had no advantages over traditional training, which had not been the subject of the research at all! Is that conclusion still in the paper?
Sorry, I was hoping to get hold of the whole article before commenting on this. Will try to read it soon. But while I have a moment to catch up on threads I thought I'd make the most of it...!
I completely agree with Sandy - this is just common sense. If the behaviour is something as simple as touching the cone then a horse is more than intelligent enough to work out the desired behaviour, with or without a click. The reason we start CT with that simple targetting exercise is to teach the click/treat association in readiness for when the behaviours become more complicated and the click is genuinely needed as a marker. Even teaching the horse to touch a row of cones would be a lot harder without a click or few.
I also agree with Lucy that generally these sort of studies don't take into account the other factors in the environment. I also worry that because of the way a probability is obtained, it doesn't allow for (for example) one horse needing the click as a marker even if the other 59 didn't. Statistically that one horse may be insignificant but if you are the proud owner of a "statistically insignificant" horse then you still need to adapt to its particular learning style rather than expecting it to behave in keeping with the other 59!!
So Lucy, Mariaana - I don't think this is something we "don't want to hear" - I think it's to be expected and is fully in keeping with our understanding with the positives of CT.
Sandy, I was really interested to hear you'd seen this paper in draft form - so do you do this sort of research then?? (and welcome to the THF!)
No Catherine, I don't do this type of research, but the article was posted -and discussed at some length! - a few months ago on the Clickryder list at Yahoo.