Bad science reporting happens on a daily basis and is perpetuated by both the media as well as the scientists themselves. UA Magazine checks whether science related media are doing justice to the research papers they mention.
Today: “Why you really can tell what your dog’s thinking: Scientists reveal we’ve evolved to identify animal emotions (which is hardly news to anyone who owns a pet)” by Daily Mail.
Are humans able to read facial expressions of dogs? Researchers took photos of a dog’s face during conditions that were expected to elicit a specific emotion in the animal, for example happiness or anger. The photos rated as best by dog experts were showed to participants – experienced and inexperienced with dogs – who tried to read the dog’s emotion.
This week, Daily Mail published an article on this research. We spoke with study leader Tina Bloom about the Daily Mail publication. Were they right?
What do you think of the title?
This study was about emotion, not cognition. In other words, we examined what dogs were feeling rather than thinking. Nevertheless, the title is not too bad, as many people tend to use thinking and feeling interchangeably.
Does this article include factual errors?
The article says “volunteers were shown a range of different images of the same dog and they were able to detect the exact emotion of the animal.” Actually, our results show that these people could NOT correctly identify disgust (on the right).
Does this article use terminology and/or draw conclusions that were not present in your study?
Yes, we were more conservative in our interpretation of the different emotional conditions than is being presented. In order to establish whether the pictures of Mal (the dog that was used) were accurately representing its different emotions, we asked a group of expert to rate them. They noted that the Jack-in-the-Box (“surprised”) condition photographs appeared to display more “curiosity”, and thought the “sad” condition was actually “submissive”. The Daily Mail article fails to mention this.
Beside the Daily Mail article, were there errors in other online publications describing your research?
Other media reported that the research was conducted in Minneapolis (where Walden is located) or Florida (where my co-author is located). In fact, the actual study took place in rural Pennsylvania, which I believe is one of the beauties of the study. Furthermore Psychology 101 failed to correctly report who particpated in our study. The population sample consisted of rural people and small town inhabitants instead of college students.
Original research paper:
Bloom, T., & Friedman, H. (2013). Classifying dogs’ (Canis familiaris) facial expressions from photographs Behavioural Processes DOI: 10.1016/j.beproc.2013.02.010