In a preliminary study on equine training, Michigan State researchers found that while adding positive reinforcement did not make horses learn a frightening task faster than horses that were handled using only traditional negative reinforcement strategies, horses exposed to positive reinforcement did not learn the task any slower, either.

Furthermore, the addition of positive reinforcement to traditional methods was considerably less fatiguing and safer for the handler.

“Negative reinforcement does not necessarily refer to something ‘bad,’ or to a punishment, but rather the removal of an unwanted consequence such as pressure on the lead rope to increase the chance of the horse performing a desired response,” explained co-author Camie Heleski, PhD, from the department of Animal Science.

In contrast, positive reinforcement involves adding a pleasant stimulus (e.g., a food treat or wither scratching) to reward the desired response or behavior.

Since horse training traditionally relies heavily on the use of negative reinforcement, Heleski and colleagues evaluated the addition of positive reinforcement to see if it would enhance learning in horses.

Thirty-four Arabians were taught to walk over a green tarp, a novel and frightening task for most horses. Half of the horses were taught to cross the tarp using negative reinforcement (i.e., pulling the halter or lead and releasing the pressure once the horse stepped forward) while the remaining horses were also trained using positive reinforcement (i.e., rewarding with a handful of grain once they stepped forward).

Upon the addition of positive reinforcement, no significant change in time to cross the tarp or time to cross the tarp calmly was observed.

According to Heleski, “More horses in the negative reinforcement group ultimately ‘failed’ the task (i.e., they would not step onto the tarp after 10 minutes of consistent efforts) compared to horses in the negative plus positive reinforcement group.”

While these results were not statistically significant, the authors suggest that the small sample size might have affected the results.

“Interestingly, half of the ‘failures’ when tested a second time with the addition of positive reinforcement were willing to cross the tarp,” said Heleski.

The study, “Evaluating the addition of positive reinforcement for learning a frightening task: a pilot study,” was published in the Journal of Applied Animal Welfare Science in July 2008.

Additional information is also available at equitationscience.com.