Verified by 
I was visiting a colleague who is associated with a dog rescue society in California. We were at the society's kennel and he was telling me about a rapid, streamlined method of temperament testing dogs for the personality dimensions of aggressiveness and fearfulness.
"It may be easier for me to simply show you how it's done," he said. "However first I have to stop by my locker and change my shirt since the one that I'm wearing has an inappropriate pattern."
I was puzzled and asked, "What does your shirt pattern have to do with temperament testing?"
"It's the stripes," he explained. "They make a dog spooky and uneasy and can bias the results of the temperament test. I'll send you a reference to the research on this if you like."
I was skeptical, to say the least, so I was pleased that by the time I returned to my hotel room, he had emailed me the reference. It was a research report that was published in the journal Anthrozoos* by Arnold Chamove of the Psychology Department at Massey University in New Zealand. His interest had apparently been captured by the concept of "aposematism" which is a form of warning coloration in animals. It is used defensively by animals that could be preyed upon in order to warn off predators. The particular patterns and colors that are associated with aposematism tell predators that this particular animal has defenses, usually in the form of being poisonous (as in the case of some snakes) or unpalatable (containing chemicals that taste bad and will make anyone who eats it sick).
Some researchers believe that the way that warning coloration works is that an animal preys upon a particular species and after eating it has a negative outcome (gets sick or hurt etc.) The suggestion is that due to that bad experience, for the rest of its life, the predator will avoid attacking that species again. Other researchers believe that this would be a fairly inefficient way for the mechanism to work. As an alternative, these scientists suggest that avoidance of particular patterns and colors in prey animals is instinctive and coded in the predator's genes.
One common aposematic pattern involves high-contrast stripes. You can see this in this poisonous coral snake.
Stripes also are a warning signal in this cinnabar moth caterpillar whose highly alkaline chemical levels make it foul-tasting and can cause nausea.
It has been suggested that the stripes on a skunk also tell predators that it will be highly unpalatable due to its glands that produce an unpleasant smelling, and highly acidic fluid.
Chamove wondered if dogs might respond to human clothing patterns which mimicked these common aposematic patterns. For the first study, 22 dogs housed in a dog pound were tested. A young male served as the model and he was dressed in black except for a long sleeve shirt that contained one of a number of different patterns. The shirt could contain alternating narrow black and white stripes (1 cm. wide) either horizontally or vertically oriented, wide stripes (4 cm.), unequally spaced stripes, or no stripes at all. This person then walked slowly past each dog's kennel, pausing for eight seconds in front of each cage door. All of the episodes, for every dog and every shirt pattern, were videotaped.
It turns out that the dogs were differentially sensitive to the clothing pattern worn by the person. The dogs were most active and emotional towards the model wearing the shirt with a narrow evenly striped pattern, and the least active when the model wore the shirt with no stripes. The horizontal stripes produced a slightly higher level of arousal in the dogs, however, the difference between horizontal and vertical stripes was relatively minor. The major emotion shown by the dogs was submission, with the animals acting as though the narrow striped shirt made them uncomfortable or anxious.
Chamove next conducted a second experiment to try to verify the results of the first study. This time, he tested 10 dogs from an animal shelter and 15 dogs from a breeding unit associated with the University. In this experiment, only three shirt patterns were used, one with black and white narrow (1 cm.) horizontal stripes (which was the most effective pattern in the first study), or a white shirt with black spots (1 cm. in diameter), or a shirt with no pattern. In this new study, the model was a female. The results were similar to the first experiment in that the narrow striped shirt produced the most submissive responses in the dogs while the plain shirt produced the least.
Taken together, these two studies seem to indicate that dogs react differently when unfamiliar people approach their enclosure wearing clothing with a striped pattern. In the main, the dogs respond with an increased level of fearfulness and a decreased level of friendliness.
As strange as it seems, these data suggest that the clothing that a person wears when they are testing the temperament of a dog can bias the results of that test. At the very least these results indicate that if a person is selecting a dog from a kennel situation they would be wise to carefully select their wardrobe so as to avoid provoking an unfavorable response from the dogs they are considering.
Copyright SC Psychological Enterprises Ltd. May not be reprinted or reposted without permission
Data from: Arnold S. Chamove (1997) Dogs Judge Books by their Covers, Anthrozoös, 10:1, 50-52
