Examining Media Myths

Why everything you've heard about how fiction influences behavior is wrong.

Posted May 18, 2020

Part I of II

When people debate media effects in college classrooms, Twitter, or elsewhere, it’s not uncommon to hear people reference historical examples of dramatic outcomes related to media.  The logic goes: “Perhaps we ought to be worried about violent video games since we know the movie Jaws led to decreased shark populations, or 13 Reasons Why led to an increase in youth suicides.”  But did these supposed events actually happen?  Is there really evidence for these beliefs or have they entered a kind of academic folklore wherein the story is just too good to undergo careful scrutiny?  In this essay, I take a look at 10 common beliefs and explore the data for each.

For the moment, I’m going to set aside comparisons between fictional media and advertisement.  These arguments typically go “If advertisers spend billions of dollars getting people to switch from Coke to Pepsi, why can’t an aggressive video game turn a kid into a mass killer?”  This is what we call in the field: a terrible question.  Advertisements and fictional media are very different.  Advertisement is designed to change behavior, whereas fictional media (some pretensions aside) usually is not.  Also, we’re talking about very different outcomes here: switching from Coke to Pepsi is a very minor behavior, though lucrative for Pepsi.  Switching from a peaceful happy youth or even a peaceful but depressed youth to a mass killer is a much more substantial change.  It’s an apples-to-oranges comparison we can entirely disregard as, frankly, a bit silly.

Comparisons of fictional media to other fictional media, though, are much more compelling.  Are there truly historical examples of fictional media having substantial impacts?  Certainly, it’s always possible.  But we have to examine the evidence for each case individually.  Just because one form of media had influence doesn’t mean “all have won and must have prizes.”  And oftentimes people’s assumptions about historical events are different from what actually happened.

As I examine these cases, I’ll refer to two levels of uncertainty.  The first outcome is what I’ll call apocryphal.  These are events people think are true, but for which data is actually lacking.  These events may be remote and so historically rooted that gathering evidence for or against the belief may be impossible.  In such cases, people asserting a relationship between a fictional event and an outcome do so without evidence.  But there’s not necessarily evidence against the belief either, so it’s still plausible on some level (though not to the degree proponents assert).  Apocryphal situations also do have one limitation we should remember: Confirmation bias.  Namely, we may focus on historical events that appear to support media effects theories, while simply ignoring myriad other events where media clearly had no relationship with attitudes or behaviors.  As such, apocryphal events may represent chance occurrences where two things appear related despite having been due to chance alone or due to underlying societal changes.  Or, put bluntly, they may be ecological fallacies, just like the observation that Nicholas Cage movies are associated with swimming pool deaths.  If we really thought Nicholas Cage was promoting drowning we might very well be highlighting this relationship as evidence for media effects!

The second class of uncertainty is debunked.  In this case, clear empirical evidence has called the belief into question.  People may still believe it but, unlike with apocryphal events where there’s some plausible deniability (albeit absent critical thinking, I’d argue), adherents to these beliefs are clearly ignoring data.  Beliefs that video games promote mass shootings certainly fall into this category by this point. 

With no further ado, let’s explore some common examples.  I’ve included some I’ve seen come up rather frequently, though this is by no means an inclusive list.  Again, for each, I’ll note whether they appear to be apocryphal or debunked. 

War of the Worlds (Debunked): We’ll start with an easy one.  Probably most people are familiar with this story: in 1938, Orson Welles broadcast his famous War of the Worlds radio show, documenting a Martian invasion of the Earth.  As the story goes, hordes of listeners believed the show to be a real news broadcast and panicked, some running into the woods to avoid the alien invasion.  This story has been hyped in news media ever since.  Only, unfortunately, scholarship finds that the fable of the panicking listeners is untrue.  There’s little documented evidence of mass panic and the vast majority of listeners correctly identified the story as a fun entertainment program.  It’s a great legend of how gullible people are to fictional media.  But it’s just not true. 

The Military Used Video Games to Teach Soldiers to Kill (Debunked): As noted above, video games have been the source of multiple moral panics (Dungeons and Dragons seems a reoccurring target as well, more below).  Do video games contribute to societal violence, sexist attitudes, or acceptance of militarism?  The answer from science seems to be largely “no.”  But video games remain a popular scapegoat from both left and right. 

A common refrain, particularly from the right, when worrying over violent video games is the claim that the US military uses violent games to teach soldiers to kill.  Typically, this claim is repeated earnestly without any data.  Tracing the origin of this particular legend takes a bit of time, so bear with me.

Back after WWII the official army historian, SLA Marshall, published a thesis suggesting that most soldiers were reluctant to fire their weapon at the enemy.  This conflicted with data from more recent conflicts suggesting much higher rates of fire and this led some people to speculate the army got better at teaching soldiers to kill.  It turns out that Marshall was probably wrong in his analysis, but whatever, the notion stuck with certain crowds.  The belief that the military uses video games to promote killing was popularized by the self-described expert of Killology David Grossman, a retired army Colonel who also apparently tells police officers they’ll have the best sex of their lives on the night after they kill a suspect.   

The military does use game-like simulators for vehicle training, team performance, and even ethical decision making.  What they decided don’t do is use video games to teach soldiers to kill.  When I, and colleagues, asked a military psychologist about this as part of a paper on drone warfare, the psychologist seemed perplexed by the question.  He pointed out the military does not want soldiers shooting everything that moves like players often do in a game, and this would clearly not be in the best interest of anybody, including the military.  Sure, if you want, you can indulge the conspiracy further and believe the military is lying, but since these claims are entirely evidence-deprived, we can consider them debunked.

Bugs Bunny and Nimrod (also Carrots) (Apocryphal): Bugs Bunny and other Loony Tunes cartoons remain the bar by which all other cartoons are measured.  Yet, in their day, some of them contained blatantly racist and other objectionable content.  They are often invoked by believers in fictional media effects.  After all, didn’t use of the word “nimrod” to refer to an idiot (as opposed to its previous Biblical reference) begin with Bugs?  And what about the whole bunnies eating carrots thing?  Carrots, it turns out, are about as good for bunnies as chocolate cake is for humans.

Let’s look at Nimrod first.  Now an insult, Nimrod began as an obscure king in the Bible.  How did the word make that transition if not for Bugs Bunny?  However, as noted in Wiktionary, some evidence suggests the use of nimrod to mean idiot predated Bugs Bunny and may have related to the Nimrod from the Bible being associated in some traditions with the Tower of Babel.  Thus, Bugs Bunny may only have been repeating the use of nimrod already coming into being in the English language at that time.  At best, proponents of the Bugs Bunny hypothesis are selecting evidence to support their views and ignoring other evidence.

What about the carrot thing?  Carrots are high sugar and can cause tooth decay and other issues for bunnies.  But they’ll eat them (hence my comparison to cake).  Did Bugs Bunny start a carrot-feeding craze for bunnies?  Here’s where it’s worth pointing out how difficult some of these questions are to answer. 

In an ideal situation, we’d have some empirical evidence for this.  For example, if we had data on the percentage of people who believed bunnies ate carrots before Bugs Bunny, as compared to the percentage of people who believed this after Bugs Bunny, particularly if such data came from a well-controlled study, that could be very interesting.  Unfortunately, we just don’t have that.  Probably some proportion of people fed carrots to bunnies before Bugs Bunny, and some people do it now.  We just don’t know if that proportion has changed.  Without empirical data, we’re left to speculate, and, scientifically speaking, anecdote-based speculation simply isn’t evidence.

It’s also an example of confirmation bias.  What I mean is, with these kinds of anecdotes people point to things that look to be correlated even if they’re not really and ignore counter-examples.  People give animals all kinds of naughty things to eat, but these aren’t based on cartoons or other fictional media.  It’s well known that ducks shouldn’t eat bread, but people feed ducks bread all the time, despite the absence of a bread-toting cartoon duck.  Then there are examples of cartoon animals with signature foods that didn’t set off a naughty animal feeding craze.  There’s no rash of people giving lasagna to cats, pizza to turtles or, for that matter spinach to sailors (and no, the rare knucklehead doesn’t prove a point).  We highlight the case that seems to fit (Bugs Bunny and carrots) and not those that don’t (Garfield and lasagna, Teenage Mutant Ninja Turtles and pizza, etc.)  This is clear confirmation bias.  Sometimes apparent relationships can pop up due to chance alone, when we focus only on these cases this falsely reinforces our belief in a correlation where none exists.

We don’t have clear empirical data one way or another.  However, with a broader view, it’s likely the Bugs Bunny thing is simple confirmation bias based in ecological fallacy.  

Jaws and Shark Populations (Debunked):  Did the movie Jaws cause people to panic about sharks, leading to extreme anti-shark views and decimation of shark populations?  Probably not.  But it’s a common belief as people address very real concerns about declining shark populations, largely due to over-fishing.  The belief became so common that even Peter Benchley expressed regret for writing the novel.

The Jaws case is a good example of how to illustrate the kind of data that would really assess whether an effect existed or not.  If we could assess people’s attitudes toward sharks before or after the movie’s release in a randomly selected sample or even do a randomized controlled trial of the movie against a control non-shark movie, we’d have some interesting evidence one way or another.  Unfortunately, we don’t.

Did people really think sharks were cuddly friends before Jaws came out?  What limited data we have suggests probably not.  Even in his statement of regret, Benchley acknowledges people’s opinions of sharks were pretty negative before Jaws came out.  As he stated “Back then, it was generally accepted that great whites were anthropophagus (they ate people) by choice” and “…people have always been terrified of sharks…”  He seems to regret falling for a popular misconception, not one he created. 

If we look at the data on shark populations, we also see that the decline due to over-fishing began years prior to the movie’s release, with the worst depopulation occurring before Jaws.  Also, to assume that hatred of sharks bounced over into commercial fishing as opposed to, say, killing them off American beaches is arguably a bit of a leap.  Further, why didn’t other “dangerous animal” horror movies such as with alligators (which just saw a new installment with the movie “Crawl”) result in massive kills (alligator populations have, if anything, dramatically rebounded in recent decades after reaching dangerously low levels)? Honestly, even sites that raise the Jaws question generally shy away from claiming causal effects (this video is a good example, slyly insinuating the question, without ever answering it or providing evidence.)  Between Benchley’s own acknowledgment that sharks were the source of paranoid fears before Jaws, and the data on shark populations declining before the movie, I think we can safely move this belief into the debunked category.    

Part I of II Posts