In somewhat of a departure to the research my group in UCL Psychiatry normally works on, I’ve just published a research paper in the BMJ’s 2014 Christmas edition on the occurrence of death featuring in children’s animated films (yes, cartoons!) compared with dramatic films for adults, jointly led with my colleague Dr Ian Colman over at the APEAL lab at the University of Ottawa.

Contains a scene young viewers may find disturbing

CARTOONS KILL: casualties in animated recreational theater in an objective observational new study of kids’ introduction to loss of life

What did we find? Well, the on-screen death of a main character, close friend or relative occurred 2.5 times more frequently in children’s animated films than dramatic films for adults. That’s quite something. Taking murder alone, this figure rose to 2.8 times excess in kids animated films, and parents were 5 times more likely to be the victim in children’s animated films compared with dramatic films for adults (perhaps this last one is less surprising, given parent’s are more likely to feature in films aimed at young children in the first place).

Hocus pocus?

By this point you might be asking yourself what could explain such shockingly high levels of death (including murder) in films aimed at young people. I hope you are! Surely this is some artefact of data manipulation, a selective bias of choosing the kids films and comparison films to load the dice to favour of this finding.

We don’t think so. Here’s a summary of how we approached the research: we took the all-time top grossing 45 animated films, indexed for inflation, and compared the amount, type and victims of on-screen death with the two top grossing dramatic films for adults in the same year of release as each of the animated films. Where we included more than one animated film from the same year of release, we included the next two grossing dramatic films for adults (i.e. 3rd & 4th) from that year.

A priori we established a few ground rules. First, we excluded sequels, since characters may have been more or less likely to have been killed off in the first film compared with the sequel (if you’re making Finding Nemo 2, you’re not going to be able to kill off Coral again). Second, we excluded animated films where the main characters were not human or animal. We couldn’t fully operationalise the concept of death in toys or machines, so we thought it would be safer to exclude those films. Third, from our comparison group of dramatic films for adults we excluded any films which received a genre tag (IMDB) of “action” or “adventure”. Why? Well many of these films are also marketed at children, think of Spiderman or Indiana Jones – and so might have biased the findings. We wanted our comparison group to be films primarily (and as far as possible) only intended for adult audiences. If you’re thinking this might have artificially excluded lots of violent movies, you can browse our full movie list here: it includes a lot of grisly movies: Pulp Fiction, The Departed, Road to Perdition, The Exorcism of Emily Rose, Black Hawk Down, Titanic.

We then watched a lot of movies. Even with a thirteen-strong team of movie raters, fantastically coordinated by Drs Mila Kingsbury and Murray Weeks – that’s an average of 10 movies each. We watched until the first occurrence of an on-screen death of a main character, or a close friend or relative of a main character and noted down the time it occurred, how it happened and who died. By “on-screen” we meant clear evidence that a character who had been present in the film had died at a certain point, even if the director spared us the most gory stills. So the shooting of Bambi’s mother, implied by her sudden absence from the film following a panicked dash away from oncoming hunters and the sound of gunshots, counts. But the orphaned princess whose parents were absent from the very start of the film would not.

We then used an epidemiological technique called survival analysis to investigate statistically whether on-screen deaths occurred more frequently in animated films for kids versus dramatic films for adults. We took into account total runtime and year of release – two potential important confounders (alternative explanations) in the study.

What lies beneath

While our findings surprised us – over twice as as many deaths of on-screen main characters or their close friends or relatives in the animated film sample – there may be a number of reasons for this. We speculated that important plot devices (like killing of the parents quickly to allow the adventure to unfold) could – when handled sensitively – also provide an opportunity for children to be exposed to difficult or complex concepts like death in a safe, warm and loving family environment. Parents might want to be on hand to answer any questions their children have while watching the films. There are also great websites out there for the concerned parent, which get underneath the certification rating of a film, to highlight any particularly gruesome, gory or otherwise potentially inappropriate content for very young audiences. Try CommonSenseMedia.org as a starting point.

We love animated films and we hope they continue to provide hours and hours of pleasure to young and old audiences alike. We hope our paper highlights some of the more surprising issues that lie beneath the seemingly innocent adventures that Bambi, Nemo, Simba and a host of princesses, dwarves and  all 101 dalmations will embark on this Christmas. That’s all folks!