By Catherine Bell, CHBC, Suzanne Rogers, CHBC, and Debbie Busby, CHBC

CHBCs Catherine Bell, Suzanne Rogers, and Debbie Busby recently published an equine welfare paper in the journal Animals. Titled “Improving the Recognition of Equine Affective States,” it represents a few years’ worth of work, completed independently of our “real” jobs. We hope it will have the effect of drawing increased attention to the subtle signs of equine fear and stress, and ultimately lead people to put their horses in stressful situations less frequently.

One of the aims of the Equine Behaviour and Training Association (EBTA) — the educational organisation that we co-founded and now run — is to “Bridge the gap between academic research and practical application.” With this in mind, we have always made sure that we have provided evidence-based information to our followers, but it is much harder to contribute to research when we are not a formal research institution. However, between us we had plenty of research, publication, and scientific publishing experience; how hard could it be? (Clue: quite hard!)

In 2015 we decided to contribute a survey-based poster presentation for the Universities Federation for Animal Welfare conference. Our abstract was accepted, and we received some positive feedback for our poster at the conference. We felt that there was potential to expand the research and attempt to publish it in a peer-reviewed journal. It was definitely worth a try.

One of the first sticking points was choosing a journal. Journals are either “open access” and funded by the authors, or funded by readers; typically in both cases it will be the institutions that pay the fees as readers are usually students. As a voluntary group with no funding, we were conflicted. We had no funds to pay for publication, but very much wanted our (largely non-academic) followers to be able to read our work. Due to a fortuitous coincidence, the journal Animals offered a fee waiver for a forthcoming special issue, and we applied, with success.

At this point we had conducted our survey and a very preliminary analysis of our data, but still needed to complete a more thorough analysis and write the actual paper. The survey consisted of six videos, each featuring a different horse that we considered — on account of his/her body language — to be undergoing a stressful experience. The videos were all different disciplines — ridden dressage, natural horsemanship, in-hand dressage, bridle-less riding, Western reining, and behavioural rehabilitation — but at no point were we judging the disciplines or handlers themselves, just the participants’ perceptions of the horses’ body language featured in each. We asked participants which of 13 different emotional states they believed the horse to be experiencing: these options were angry, anxious, conflicted (defined in the survey as “experiencing two emotions at the same moment in time”), enjoying it, excited, fearful, frustrated, playful, relaxed, stressed, stubborn, submissive, and switched off (or “resigned”). We also asked whether participants would be happy to see their own horses experiencing the handling, and gathered various pieces of demographic information such as preferred discipline and equestrian experience.

Our experiences of working with clients had taught us that recognising the subtle signs of fear and stress is a rare skill, and we hypothesised that the data would confirm this. As is so often the case, things were not quite so straightforward. Our participants (opportunistically recruited via Facebook, so not a representative sample of the equestrian industry, just the best we could do at the time) seemed to be very good at recognising the body language in some of the videos, but not all. Our statistical analysis found that the videos featuring natural horsemanship and bridle-less riding were particularly misinterpreted as positive experiences for the horses.

This is something we commonly see in our consulting as well: Euphemisms and flowery language result in many people believing that such training practices are benign, with the horses described as “playing” and engaging with the handler in a manner analogous to equine herd behaviour. Of course, a little more consideration results in the recognition that the training at best invokes negative reinforcement and punishment, and often stretches to flooding. This results in some very stressed horses whose body language is mistaken for “relaxed.”

The subtle — and not so subtle — signs of stress can be seen clearly in this horse’s facial expression.

Another interesting finding was that increased owner experience did not lead to increased likelihood of interpreting the body language correctly. Even working with horses professionally was no guarantee that the participant could recognise equine fear or stress. Deep down we are probably not surprised by this — there are plenty of horse professionals whose empathy for those horses is lacking. But it is sad to see the numbers reflect this, particularly when some of those professionals indeed recognised the stressed body language but considered it acceptable to continue putting the horse in such a situation.

When horses are trained with shaping and given the opportunity to take their time to explore their environment, their experiences can be much more positive.

The only category of participant that did appear to suggest an increased likelihood of recognising equine fear and stress were the clicker trainers. Now, it is tempting for clicker trainers to become excited at this result, but we really need to exercise caution in the interpretation. The proportion of clicker trainers in our overall sample was small and the basic statistics were suggestive but unconfirmed by the more complicated statistical analysis. Having said that, we would expect clicker trainers to be more able to read equine fear and stress, because the whole point of the training is to ensure that the horses have a positive experience. With that in mind we should be concerned that it was far from 100% of our clicker training participants who gave correct answers, and we encourage clicker trainers to avoid complacency in this respect! We would love to do some follow-up research here — watch this space…

As we wrote the text of the paper we were very conscious of the language we used. Despite this, the peer review process identified many areas of the text where it was felt we had been sloppy in our language — particularly the use of “emotion” instead of “affective state,” and “stress” instead of “distress.” To some extent this may have been reflective of our editor and reviewers, as our language often seemed (to us) comparable to that of other papers. But our previous experience of research is that it is better to “pick your battles” during the peer review process and concede as much as possible with good grace.
Another key lesson we learnt during the peer review process is that, back when we had designed our survey, we really should have obtained approval from an ethics committee. EBTA doesn’t have such a thing, and we are still not sure what to do about it for next time, other than have collaborators who are members of academic research departments. Thankfully we had been sufficiently aware of good ethics practice and followed the British Psychological Society’s guidelines sufficiently well that we were able to be accepted for publication without official approval from an ethics committee.

Conducting this work has been a great experience with a steep learning curve. It is something we would very much like to do again, and we have a number of ideas in the pipeline. In the meantime, the full text of our paper is freely available to all at https://doi.org/10.3390/ani9121124.

Catherine is a CHBC in the South-East of England. In between home-educating her two children, she works as an equine behaviorist and independent barefoot hoof trimmer. She hosts the Thinking Horsemanship Forum and is a co-founder of the Equine Behavior and Training Association.