Association between violent video games and aggression among adolescents
psycnet.apa.org
psycnet.apa.org
Another trash psychology study that takes a correlation and tries to claim it's a cause-effect relationship.
> In the past 2 decades, correlational and experimental studies have found a positive association between violent video game play and aggression.
"Correlational ... positive association ..." Indeed. Everything but the words "one caused the other."
"Sustained violent video game play was significantly related to steeper increases in adolescents' trajectory of aggressive behavior over time."
Yes, and still a meaningless correlation. Did the video game cause the aggressive behavior, or the reverse? Or is there some third factor that caused both? We don't know and this study can't tell us. I am perpetually amazed at what passes for science among psychologists.
"Our findings, and the fact that many adolescents play video games for several hours every day, underscore the need for a greater understanding of the long-term relation between violent video games and aggression, as well as the specific game characteristics (e.g., violent content, competition, pace of action) that may be responsible for this association."
Here, by trying to claim a cause-effect relationship but one not based in evidence, the psychologists cross the line into pseudoscience. There is precisely zero scientific evidence on which to base a claim that there is a cause-effect relationship between aggressive behavior and video game play (or the reverse). All we can say for sure is this is what the researchers hoped to find.
A real scientific study would empanel two groups of teenagers as identical as possible, force one group to play video games, and forbid this activity to the other group. About that proposal I can say three things:
1. It would qualify as science in a field with copious amounts of pseudoscience.
2. It would be unethical in the extreme.
3. It will never happen.
This study is a perfect picture of "psychological science", which is to science what military music is to music.
I ask because I only was able to get the abstract, so there might be something in the rest of the paper for your arguments. However, the abstract seemed very careful not to make overly broad conclusion. It says the playing violent videogames is a good predictor for being agressive, and being agressive is not a good predictor of playing violent video games.
This is not sufficient to claim causation, which they do not (in the abstract at least), however this type of corralation that exists only where A preceeds B definantly suggests it.
As to you concerns about sampling, they seem to be following best practices for this type of study, random. As long as both groups you are comparing are selected using the same random, your results are valid (for the group which they were selected from). Most statistical tests (which I assume were used in the main paper) are designed assuming the groups are random, and will provide valid statistics.
In general, but when they say "... long-term relation between violent video games and aggression, as well as the specific game characteristics (e.g., violent content, competition, pace of action) that may be responsible for this association", they're moving from evidence to opinion, an opinion not remotely supported by the study.
The only way they could sort out causes and effects in a study like this would be to compel some people to play video games, and compel other people not to. The subjects are pre-teenagers, and that is not going to happen. Such a study is called "prospective", and they are rare in psychology. Even with adults, prospective designs are terribly expensive -- and there have never been a study of that kind on this topic.
> As to you concerns about sampling, they seem to be following best practices for this type of study, random.
But the study has no control. Instead of choosing two groups as identical as possible and applying different stimuli to each, they rank subjects based on their existing behavior, meaning there is no control group as that term is understood.
> Most statistical tests (which I assume were used in the main paper) are designed assuming the groups are random, and will provide valid statistics.
But there aren't two groups selected in advance and randomly assigned a behavior. The study participants are pigenholed based on their observed behavior, the reverse of the strict statistical approach.
On can see this without knowing the details, simply by imagining the opposite -- locking up a bunch of 12-year-olds and compelling them to play, or not play, video games. Obviously that didn't happen.
> As long as both groups you are comparing are selected using the same random ...
There weren't two groups selected in advance and subjected to different conditions, as in a truly scientific study. This study design is called "retrospective" (meaning subjects are chosen after the fact, not before), and is notoriously unreliable.
http://en.wikipedia.org/wiki/Retrospective_cohort_study
A quote: "Caution needs to be exercised in particular with retrospective cohort studies because errors due to confounding and bias are more common in retrospective studies than in prospective studies."
violent games -> aggressive behaviour
aggressive behaviour -> violent games.
here for example the socialization-hypothesis. So it seems, that they have found a way to debunk the theory, that more violent teens tend to play more violent videogames. they controlled for pre-playing aggression-levels, and they monitored the developement of aggression over a longer period of time, to see, if the level of aggression changed (as it seems to do), when playing violent games.
So yes, the can only show a more or less strong correlation between these two factors (and there might be a lot of other factors not controlled for), but they can show, that some other hypothesis does not explain the data.
not more - but not less either.
A non-causal association would have to be induced by a confounding factor that first causes teens to play violent video games then to behave aggressively, but never in the other order.
The causal effect is reinforced by the dose effect in the context of this temporal sequence, i.e. the more hours they play per day, the more likely they are to behave aggressively.
Yes, but not "never in the other order": just more often/strongly in the games-first order.
And I don't find such confounding factors hard to think up for 9th-12th graders: difficult family life; social problems with peers; substance abuse; academic performance pressures; etc.
If a teen were to react to any of these, first, by escaping into violent video games (as a relatively easy temporary mood boost), but then second, with measurable aggression, it would fit the sequential relation seen. But we wouldn't know with any sort of confidence if the violent-gameplaying interim behavior worsened or lessened the eventual aggression.
"Sustained violent video game play was significantly related to steeper increases in adolescents' trajectory of aggressive behavior over time. Moreover, greater violent video game play predicted higher levels of aggression over time, after controlling for previous levels of aggression, supporting the socialization hypothesis. In contrast, no support was found for the selection hypothesis. Nonviolent video game play also did not predict higher levels of aggressive behavior over time."
They write that they find support for the "socialization hypothesis", presumably meaning that players pick up norms from the violent games.
They define socialization/selection in the abstract:
the socialization (violent video game play predicts aggression over time)
versus selection hypotheses (aggression predicts violent video game play over time)Perhaps, some other childhood stressor or failure in coping mechanisms? And if this other factor tends to trigger first, escapism into violent gameplaying, but then later (if unaddressed), actual aggression, I think they could see the "steeper increases in adolescents' trajectory of aggressive behavior" that's reported, even though the gameplaying is just a waypoint or signal rather than cause.
(Under this hypothesis, it's possible the violent gameplaying aggravates the underlying issue, if it delays or prevents other remedies, but it's also possible the gameplaying serves to soften the aggression. You can't really tell from their sequenced relations analysis... you'd want some stronger random-like control on the amount of violent gameplaying.)
From Abstract: "and a comprehensive set of potential 3rd variables were included as covariates in each analysis".
Sounds like they certainly tried to control for that - you'd have to read the full study to see the details.
I recall many aspects of my and my peer's psychological lives that would not be easily available to an outside researcher as part of their "comprehensive set of potential 3rd variables".
I remember as a kid we played war games outdoors, i think kids have always done that in some ways.
When I was a kid (70s) our "war games" were quite remarkably violent - I'm surprised that nobody got killed.
These days I watch my teenage son play rugby - which is itself rather violent (although generally very good natured) and from what I can see it is a pretty good sport for teenagers who like playing it (as my son is going to a rugby-oriented school he is lucky, it must be miserable for kids who hate rugby).
Yes, they say that, but the study offers no evidentiary support for that opinion. Scientific papers are not supposed to be soapboxes for the opinions of the researchers, only the careful reporting of scientific results. And the result of this study is that there is a correlation between aggressive behavior and video game play. Correlation is not causation. Aggressive behavior might lead to a preference of video games, or the reverse. This study cannot sort it out. And a study that did sort it out would do it by forcing people to play, or not play, video games, for an extended period. That's unethical and will never happen.
You speak from experience? There is strong orthogonal logic that supports the hypothesis that simulated violence is effective in generating, inducing, and manipulating aggression in human subjects.
> You speak from experience?
It's not as though I invented this idea:
http://en.wikipedia.org/wiki/Correlation_does_not_imply_caus...
> There is strong orthogonal logic that supports the hypothesis that simulated violence is effective in generating, inducing, and manipulating aggression in human subjects.
There is no evidence whatsoever to support a cause-effect relationship such as you suggest. If a study were to be designed that could uncover such a relationship, it would violate experimental ethics and would not be funded.
And your link offers no scientific support whatsoever for your claim. It's an opinion piece, not a scientific study. It offers precisely zero evidence in support of its opinions.
By your definition, Vietnam was a scientific experiment? This speaks volumes about how you picture science. The plural of anecdote is not evidence.
Data (roughly) for effectivenes / lethal intent:
Population (normal): 5%
Soldier (WWII era): 20%
Soldier (Current Era): 95%
These are anectdotally referenced in the linked article, but they are discussed in that article's main citation, which is here:
Publication Date: November 1, 1996 | ISBN-10: 0316330116 | ISBN-13: 978-0316330114 | Edition: 1
[Amazon.com] Lt. Col. Dave Grossman draws unsettling, even sinister parallels between the psychological conditioning required to make soldiers kill in war and the similar effect that videos, films, games and movies have in civilian society.
This book is/was a textbook used in the armed forces, it is a summary of working papers etc. It is not some BS handwavey claptrap from academia. What you will notice from reading it is it is actually very difficult to get people to Kill one another.
That is why the data on effectiveness are so damning.
And you are perfectly scientifically illiterate.
> In preparation for vietnam, the US armed forces studied lethality of intent for small arms fire.
Yes, they did, But it was not a scientific study -- no control group. No basis for comparing results. No basis for falsification. As a result, they "discovered" exactly what they expected to discover.
The entire Vietnam episode was based on the opinions of experts, not scientific study. And I am hardly the only one to make this observation.
> This book is/was a textbook used in the armed forces, it is a summary of working papers etc.
Yes, and it is not science. I now realize I'm talking to someone who doesn't understand what makes science science. And that is ... drum roll ... falsifiable theoretical claims, claims that someone could in principle conclusively falsify in practical tests. There are none in the book you cite -- it's a sequence of anecdotes and philosophical speculations.
> It is not some BS handwavey claptrap from academia.
I agree with the "from academia" part -- but it is certainly contentless philosophical speculation rather than science. There are no testable, falsifiable claims between its covers.
Strawman. Sorry. Your understanding of human behaviour is facile. The purpose of orthoganal explanation is to avoid this. There is not stronger form of argument, please don't pretend to lecture.
On the data, you're argument is from Ignorace. The author of that book is one of the most knowledgeable people on the planet on that subject. Like it or not, your opinion has not been substantiated by any form of counter argument.
I don't need "objective" argument when I have intersubjective authority. You're argument about "science" is laughable in its in-applicability and ultimate irrelevance.
Which is obvious as well from your trivial citation of correlation/causation.
You don't understand how to create an argument and verify it using empirical data. The practicum on this is evidenced by the techniques of Special Forces operatives and all kinds of actual professional killers. These people don't have the inclination, the incentive, or the bandwidth to develop techniques which are failures.
The fact that you think Academics hold a superior position to actual killers on the subject points to the double weakness in your argument: (1) Academics have no actual experience; and (2) academics have all kinds of conflicted incentives to lie, the least of which is there pre-held political beliefes and naive understanding of actual human nature.
Try again. Or actually, don't. Actually, just read that book. The link to this discussion was not the motivation or the functional purpose of that series of working papers. The latter just became obvious and self/evident later (much of the background material, for example pre-dates the mid-1980s).[1]
____________
[1] Edit: As an aside, where do you think Andres Breivek got his idea to prepare for the massacre on Utoya? You should do some research on that as well. Both before and after there is relavent data on the record about it. The data there are also in distinct contradiction to your "argument" here.
> Strawman.
You need to look up the term "straw man." My mentioning that the source you quote is not science is not a straw man, because that's the topic under discussion.
> Your understanding of human behaviour is facile.
Straw man. The topic of discussion is not human behavior, but whether or not psychology is a science. Or have you forgotten?
> There is not stronger form of argument, please don't pretend to lecture.
To someone who doesn't understand science, but who presumes to pontificate on the topic? One who thinks a philosophy textbook is science?
> The fact that you think Academics hold a superior position to actual killers
What the fuck are you taking about? You locate where I ever said or implied this, anywhere, and do it now. I happen to hold the opposite view, but very clearly, evidence is not a matter of concern to you.
> The author of that book is one of the most knowledgeable people on the planet on that subject.
And Hemingway is an authority on bullfighting -- but that doesn't make his books scientific.
Honest to God. I give up -- your ignorance is too profound.
There's no sense in being precise when you don't even know what you're talking about
> That about sums it up. Also, its clear you haven't read the source material you reference (ie, the source not the summary).
You'd be surprised about what people know. And what goes on behind closed doors.
Not sure where you get your expertise from...
I hope its not just playing video games/.
Actually, it seems to make perfect sense. [1]
_______
[1] He's a rocket scientist.... Who I have a lot of respect for, but this isn't rocket science (excuse the pun). The arguments here are out of place. There is a time and a place for applying (correctly) different frameworks of logic. This sub-thread was explaining the existence of <orthogonal> support, viz: http://news.ycombinator.com/item?id=4626135.
12-year-olds? The study we're discussing examines the behavior of minors. Minors cannot give the kind of consent you're suggesting.
Also, the sort of strictly designed experiments that might turn psychology into a science, are extremely expensive and not likely to be funded. The reason? Research standards in psychology are so abysmally low that a well-designed study would be so expensive by comparison that the granting agencies would refuse to fund it.
Have you actually read the study in full? If your claiming that the study does not in fact prove what the abstract says it does, then you'll need to provide some more details on that.
The scientific burden is not mine to show that the article reports a correlation and nothing more, the burden is on the authors to show that that they have demonstrated a cause-effect relationship. And based on their study design, they wouldn't dare -- they have no way to distinguish causes from effects, as you would know if you actually understood the article and the study it reports.
I understand that you want parents to be informed about their choices for their children, but unfortunately, this kind of study has a history of being (ab)used by groups trying to limit adult's access to this kind of video game, which I have a serious problem with.
Let's not "throw science under the bus" at all. In fact, there have been studies[1] which show violent crime decreases as access to violent video games increase.
[1]: http://www.telegraph.co.uk/technology/video-games/8798927/Vi...
The study linked above looks at mid/long term effects on a large-ish cohort in a real world setting.
> The study in the article you link to shows that violence decreases in the short term when a violent block buster is released, because violent individuals are busy playing.
That's right, but I fail to see how that's a flaw in the study?
[1]: http://content.usatoday.com/communities/gamehunters/post/201...
First, a few points:
* Many things are "not meant" for teenagers, yet teens acquire (or do) them anyways. Examples include: alcohol, tobacco, watching R-rated movies, etc.
* You may have forgotten just how old kids in highschool are. Specifically, these are teenagers from 14-15 (Grade 9) to 17-18 (Grade 12). While no one under 18 should be playing M-rated games, these individuals are hardly babies.
To answer your bigger question: studies like this are important because teens are a huge video game demographic and are still developing (mentally and physically) during their highschool years. Only now are we seeing such an abundance (and easy access to) video games of all kinds, including violent video games. We must have more such long-term studies in order to understand any potential risks for Jimmy if he spends 3+ hours every day playing a game in which his character "kills" humans (who are mostly foreigners and often in gruesome ways).
You are welcome to mention parental responsibility and voluntary rating guidelines (and these are good points). I'm also against the terrible, blanket attempts at regulating violent video games we've seen in California and other states. However if we discover a link between violent video games and long-term increases in teen aggression then I see nothing wrong with strengthening legislation concerning the sale of M-rated games to minors.
That's the main weakness in the study, btw. It could well be that kids who are prone to violence are more attracted to violent games.
Given the popularity of video games in American life it just seems odd that we haven't seen a massive spike in crime given the number of mass murder simulators we've been exposing our young'uns to since arcades became popular in the late 70's.
That's not to say that immersive violent video games don't have any effect - but given how much young males play video games (they also happen to commit the most crimes http://bjs.ojp.usdoj.gov/content/homicide/ageracesex.cfm ) it seems strange there's no visible uptick - if the sky is falling, wouldn't there be some way to tell?