Some dogs bark when their owner is out and they are left home alone. A recent study by Alexandra Protopopova (Texas Tech University) et al investigates the effectiveness of a humane, automated approach to solving barking problems.
The research was conducted because some owners use citronella or shock collars to try and prevent their dogs from barking. While the devices may sometimes work, there are concerns they may also have adverse effects.
For example, if a dog barks when they see people going by the window and then receives a burst of citronella or an electric shock, they may associate the unpleasant experience with people and become fearful and/or aggressive. Because of these welfare concerns, some organizations recommend against their use (see the AVSAB position statement on the use of punishment).
This study used a humane approach that rewarded dogs with food (via a PetSafe remote-activated feeder) for periods of quiet. It was not a fully-automated system as the researcher logged barks and activated a remote control, but it shows the possibility of an automated system in future.
Eight dogs were initially recruited to take part in the study, but three were almost immediately excluded when they failed to bark during the first two sessions. The remaining five dogs ranged in age from 8 months to 6 years.
Rewarding the dog for being quiet is what is known as a DRO – differential reinforcement of other behaviour, i.e. other to barking.
The period of time dogs were expected to be quiet for was different for each dog, based on observations of the frequency of barking. For two of the dogs, it was as little as 5 seconds, and for another it was 7 seconds. For these three dogs, each session was only 10 minutes long so that they did not eat too much; the other two dogs had 20 minute sessions.
The design of the study involved a baseline period in which the dog is left alone and barking is monitored but nothing happens, followed by a test period in which the feeding system was used to reward periods of quiet, and then a repeat of both sessions.
The owner left the dog, either shutting the dog in a room or crate as they usually did when leaving home. The researcher was in another room where they could hear barking and activate the remote control when the software told them it was time to give a treat.
During the test sessions, the interval for which each dog was required to be quiet before getting a treat remained the same for the entire time.
For example, a dog called Nina barked every 4.4 seconds on average during the two baseline periods. During the test sessions, every time she went 5 seconds without barking, she was given a treat. During these 10 minute sessions, she barked on average every 26.6 seconds – but in fact she did not bark at all during the second test session.
The protocol worked for three of the five dogs. It did not work for one dog, and for the other dog it was not easy to tell.
In a second experiment with just the three dogs for whom it worked, the length of time they had to be quiet before earning a treat was increased by doubling the time from one session to the next. Two of the dogs were quiet for the longest period tested (600 seconds and 1,200 seconds). The third dog (Nina) showed big improvements in the early stages but then began to sometimes bark again; this may or may not have been related to a mistake that increased the duration more rapidly than planned.
One thing to note is that the dogs could have been barking for any reason to be accepted into the study. In fact the paper says four of the dogs (all except for Nina) potentially showed signs of separation anxiety.
The time intervals were chosen based on what it seemed a dog would be able to easily achieve, based on their barking record, but the smallest time interval used was 5 seconds. Maybe a shorter interval would have been better for some dogs. Also, it would be more efficient to have a protocol for adjusting the time intervals throughout, but that was not part of this study.
The scientists conclude “This study provides evidence of the efficacy of an alternative (DRO) to the devices that deliver aversive stimulation to decrease home alone excessive barking for at least some dogs.”
Think about it: just two 10- or 20 minute sessions were enough to teach three of the dogs to be quiet for a certain length of time. Further research can investigate the best training protocol to use, and the most efficient way to increase the duration of the quiet behaviour.
This study has a very small sample size, but the results are promising. Perhaps in the future there will be a range of automated reward-based anti-barking devices on the market.
People who buy citronella or electronic anti-bark collars might be just as willing to buy an automated anti-bark feeder and use it to provide their dog’s meals, giving them an easy solution without welfare concerns. It is already possible to buy such a device (although not the version used in the study which was designed by the researchers).
A good reward-based dog trainer will devise a training plan that takes into account the reason why the dog is barking (e.g. excitement, fear, separation anxiety). If you think your dog has separation anxiety, you will find useful information via the website of Malena DeMartini.
Does your dog bark a lot?
Companion Animal Psychology is a participant in the Amazon Services LLC Associates Program, an affiliate advertising program designed to provide a means for sites to earn advertising fees by advertising and linking to Amazon.com
Protopopova, A., Kisten, D., & Wynne, C. (2016). Evaluating a humane alternative to the bark collar: Automated differential reinforcement of not barking in a home-alone setting Journal of Applied Behavior Analysis DOI: 10.1002/jaba.334
Photos: Susan Schmitz (top) and Jennay Hitesman (Shutterstock.com)