From the results in this study it can be concluded that the horses did not have a preference for their owner (or the stranger for that matter) in either of the tests. This lack of preference has also been observed in other similar studies.
In this study, the lack of preference could be because the tests may not actually have been about the horses’ preference as I had planned them to be but rather about the horses’ curiosity and exploratory behavior. Perhaps the horses did not show a preference due to the fact that they were loose in the test arena during both test and were allowed to explore on their own. In the odor preference test it could also have been about novel objects as the buckets regardless of odor were unfamiliar to the horses and that they therefore might have explored both buckets somewhat equally.
Another explanation for the lack of preference could be that the attachment bond that is expected to be observed is similar to the bond between a dog and its owner when it in actuality might be quite different. In that regard it may be more similar to the attachment bonds observed between cows and humans where it is less about an attachment bond to a specific human and seeking out that particular human in a stressful situation and more about humans in general serving as safe haven during stressful situations.
This could also be why the horses with intermediate and high stress scores sought more human contact in general regardless of if it was their owner or a stranger than horses with low stress scores.
Why type of horse also had an effect on behavior could be because studies have previously observed that different kinds of breeds differed in their behavior. I did not include specific breeds in this study (except the Icelandic horses), however behavioral differences between breeds could still explain the differences I observed between full-size horses and ponies as well as Icelandic horses as some full-size horse breeds have been observed to differ in behavior from some pony breeds.