Cats run the Internet, so it’s not surprising that stories reporting on the latest feline antics proliferate like litters of feral kittens. Research on our furry friends is catnip to media organizations, which often write about subjects they find interesting, cute, or funny — and then those concepts become accepted as facts. Unfortunately, news organizations and blogs don’t always report entirely accurately on scientific research, choosing to report the “punchline” but not all the facts behind it. Take, for example, the recent breaking news that tortoiseshell cats really do have serious attitude, confirming a belief that dates back to at least 1895, when a veterinarian commented that torties are: “not [overly] affectionate, and sometimes even sinister and most ill-tempered in [their] disposition.”
Fact or fiction? “Well … we need more data,” is the short answer, but explaining the reasons illustrates why you shouldn’t believe every cat study you read about.
First, we’ll address the story at hand. Here’s an example of a headline you might have spotted: “UC Davis Study: Calico, Tortoiseshell Female Cats Often Most Challenging.” The Sacramento Bee was among publications that reported on a study claiming that these infamously zesty cats are indeed peppier than cats of other colors, confirming what many people already believe. That right there should raise a big question. What follows is a reader’s guide to picking science reporting apart.
When considering a study, who, what, when, where, why, and how are all critical questions. First, though, you’ll need to find the study, and some mainstream journalists fail to cite or link to that, forcing you to hunt it down. The tortie research appeared in the Journal of Applied Animal Welfare Science, Volume 19, Issue 1. It has something called a Digital Object Identifier — a code you can use to locate it anywhere, even if links expire, because it’s uniquely tagged to this specific study. That number is 10.1080/10888705.2015.1081820, but I’ll link you directly. Now for the relevant questions.
Anyone can conduct a study. See whether the science was conducted by reputable researchers working under the auspices of a rigorous organization. The University of California, Davis is renowned for its veterinary science and agriculture programs, so it’s likely that that researchers know what they’re doing. If the research affiliations don’t look familiar, look them up. Pay especially close attention to companies and foundations, because they can betray hidden motivations. A study concluding that a given insecticide is safe to use around cats, for example, is worth nothing if it’s sponsored by the company that manufactures it.
Standards for academic journals also vary. Some have extremely liberal acceptance standards and don’t review submitted work too rigorously. Others might scrutinize every submission and require multiple rounds of revisions. Older journals tend to be more well established and choosy about research, suggesting that their publications carry more weight. The Journal of American Medicine, Science, and The Lancet are some examples. The Journal of Animal Welfare Science has been in publication since 1998, and you can learn a lot about its reputation within the field from Elsevier, which provides quick overviews profiling academic journals along with their rankings along a number of criteria based on citations from other journals. Because the journal is relatively young and it fills a specialist niche, view those rankings with some skepticism — it’s competing in a small group, which means few other journals would be referencing it in the first place.
Thanks to the standardized nature of scientific publishing, the answer to this question is defined in what’s called the abstract. In this case: “The authors explored a possible relationship between coat color and aggressive behaviors in the domestic cat.” One sentence tells you what they considered and which population they studied. The next sentence in the abstract discusses a survey, alerting you to the fact that the study isn’t something like a literature review, where researchers look at dozens of studies on a related subject to get a larger picture, nor was it an experiment involving numerous cats of different coat color.
This can vary in usefulness depending on the subject. With a survey, it helps to know that current cat guardianship was concurrent with the survey — in other words, people weren’t asked to reflect on cats from their past. You’ll find out more about that when you read the survey methodology. In another example, researchers might look at long-term outcomes for cats with compromised kidneys. If the study is dated 1996, it’s not as useful as a study from 2006 or 2016, because veterinary medicine wasn’t as advanced in 1996. Basing health care decisions on old research isn’t a good idea.
Was it done in the United States? Abroad? What kinds of prevalent social attitudes might have influenced the researchers or their subjects? Did the research take place in homes or in a lab? These factors can all make a huge difference in study outcomes. In this case, the researchers used Facebook and two listservs (Cats and Kittens and For the Love of Cats) to recruit participants. Most were from the U.S., but respondents also came from Australia, China, and places in Africa. Using social media to distribute links to surveys is common, but it comes with a big flaw, one the researchers themselves acknowledged: It creates a self-selecting convenience sample, rather than a truly random one. The people most likely to complete a survey on cat behavior may not necessarily provide a balanced view of how cats behave.
What was the initial goal? Researchers can fall into a trap known as confirmation bias — they’re looking for information that will confirm what they think they already know. Some good reasons to conduct research could include: exploring social attitudes, gathering information that might be useful for veterinarians and shelter workers (whether in their practices or in communicating with the public), or developing new understanding about cats in general.
The objective of this research was to determine whether cats of any particular coat-color pattern are more likely to engage in the following: aggressive behaviors toward humans, aggression toward people during handling (punishing, petting, or grooming), or aggression during veterinary visits. While we were interested in exploring any relationship between coat color and aggression toward humans, we hypothesized that sex-linked females (tortoiseshells, calicos, and others) would be represented significantly more often than any other color pattern as being aggressive toward people.
This could be the most important question. Study methodology is critical, because it helps the reader determine the validity of a study — and lots of journalists skip over it. As discussed above, the research was a survey, and the researchers discuss the flaws of this method in their conclusions. They also note that responses were subjective — opening up the possibility of confirmation bias or a self-fulfilling prophecy — and that some rare coat colors weren’t well represented. Researchers also acknowledged that genetics (coat color) isn’t the only contributing factor to aggression when issues such as home environment, neuter status, and breed may play roles as well. The researchers also expressed some regret over the methodological choice of measuring handling aggression, aggression, and veterinary aggression slightly differently, making it difficult to run a comparative analysis.
Beyond these specifics, another issue often gets left out: study details. For example, overall, the study found that female cats tend to be aggressive, something left unreported in many stories. What’s more, white-and-gray cats rank near tortoiseshells and calicos on the aggression scale. The researchers also establish a tentative link between black-and-white cats and aggression. The researchers were surprised by the news on aggression in white-and-gray and black-and-white cats, which is evidence that they did their jobs: Rather than neatly confirming their hypothesis, they expanded our understanding of possible links between coat color and aggression, and they illustrated the need for more research.
However, the researchers also made a very important point: The variation in aggression between coat colors was fairly low. They suggested that this might be attributable to cats in general being pretty calm animals, or that guardians might be viewing the survey questions slightly differently (another example of how survey respondent issues can create study bias). For example, one might not consider a gentle warning bite an act of aggression, while another would mark it down as such.
Finally, vets should be pleased: Researchers found no significant variations in aggression level associated with coat color at vets’ offices, despite common attitudes among veterinarians, technicians, and shelter workers that some coat colors reflect cats with more peppery personalities.
To recap: We started with “Calico, Tortoiseshell Female Cats Often Most Challenging.” But based on what we learned about the study, a more accurate summation might be “Certain People on Social Media Who Saw a Cat Survey Reported That Their Female Tortoiseshell, Calico, or Gray-and-White Cats Show Slightly More Aggressive Behavior Than Other Cats.”
Top photo: Tish, LuAnn Snawder
About the author: s.e. smith is a cat-owned writer, editor, and agitator living in Northern California with felines Loki and Leila. While not mediating cat fights, s.e. explores a wide variety of subjects in writing and elsewhere, in addition to enjoying reading like a fiend and baking like an angel. Follow smith on Instagram, Facebook, and Twitter.