I'd be curious to hear why you think that any single cause of violent behavior,
which is obviously multiply determined (as is virtually everything we study),
would account for more than 10% of the variance. And as far as the "less than
10% of the variance" criticism, I'll adapt an example from Rosenthal and
Rosnow: imagine a design in which a researcher compared kids who played violent
vs. nonviolent video games to see whether they became school shooters, and
found the following results:
nonviolent games
violent games
did not become shooters 66 34
became shooters 34 66
What percent of the variance do you thnk playing video games accounted for~ If
you said 9%, you're right (r = .30). Is this a trivial effect~
________________________________
From: Don Allen [EMAIL PROTECTED]
Sent: Thursday, March 27, 2008 5:11 PM
To: Teaching in the Psychological Sciences (TIPS)
Subject: Re: [tips] Computer games to get cigarette-style health warnings -
Times Online
Hi Mark-
I just re-read the Carnagey and Anderson paper and I'm still not impressed. In
the Method section re Exp. 2 they say, " The participant also rated the video
game on various dimensions (difficult,absorbing,action-packed, arousing,
boring, enjoyable, entertaining, exciting, frustrating, fun, involving,
stimulating, violent, and addicting)." However in the Results section they
merely say, "Note that the effect of violence was obtained even though the
violent and nonviolent games were equally arousing and all games were
competitive." without providing any data whatever. A strange omission don't you
think? Then in Exp. 3 they say, "Also,several video-game ratings (absorbing,
boring, enjoyable, entertaining, exciting, fun, involving, stimulating,
addicting) predicted aggressive behavior, Fs(1, 134) = 4.75, ps <.05". Doesn't
that just make my case? Even if this one study did stand up it has not (to my
knowledge) been replicated by an independent lab. The studies on the Mozart
Effect looked good on their own, but replication proved to be a problem.
As far as the Anderson and Bushman study; it's just the old "bundle of sticks"
argument. Each study may be too weak on its own to prove the case, but if we
take them all together then they must constitute a proof. As far as I'm
concerned a bunch of weak studies is just that; a bunch of weak studies.
More importantly, even if all of Anderson's assertions were true he is still
only accounting for less than 10% of the variance. If you are really concerned
about violence then focus on important issues like economic disparity and
prevelence of handguns.
I still remain in the skeptics corner.
-Don.
Mark A. Casteel wrote:
Hi Don. For a good study that (in my opinion) satisfies your requirement for a
high-action no-violence group, take a look at Carnagey and Anderson (2005). Ps
played either a violent version of Carmageddon 2, a version where violence was
punished (points were lost for killing people/hitting objects, and a
no-violence version (same game) where violence wasn't possible. Aggressive
affect, cognition, and behavior all differed in the reward compared to the
no-violence group, and aggressive cognition and behavior differed between the
reward and the punishment groups.
Also, what about the meta-analysis done by Anderson and Bushman (2001), which
found no gender diffs and no diffs as a function of experimental vs.
correlation studies? Granted, the rs were all in the range of .16-.27, but as
the authors note, the average effect size was the same as that between condom
use and the prevention of HIV. Do we take that association seriously?
-- Mark
At 02:06 PM 3/27/2008, you wrote:
It's this kind of junk science that really gets my goat. It is highly
reminiscent of Frederick Wertham's "Seduction of the Innocent" which set off a
frenzy for banning comic books because they were corrupting our children. For
those of you who are too young to remember this era you can find a good
discussion of it here:
http://art-bin.com/art/awertham.html
I have reviewed the literature that purports to show a causal link between
media violence and violent behaviour in children and I remain highly
unimpressed. Most of the studies are correlational and even there the
correlations are weak (.20 to .30). The studies that actually attempt to
manipulate exposure to violent media are all badly flawed. They merely compare
children who watched a violent video (or played a violent video game) with
children who watched a non-violent version. What's wrong with that? Well, for
one thing the violent media were also action-packed while the non-violent
comparitors were dull as dishwater. Since violence and action are confounded in
these studies you need a third group (high action-no violence) to determine
which component produces the effect. So far, I have been unable to find a
properly controlled replicated studies that clearly establishes a causal link.
Until I see one I'll remain in the skeptics corner.
-Don.
Christopher D. Green wrote:
Check out this Times (of London) article on British efforts to put warning
labels on video games.
http://www.timesonline.co.uk/tol/news/politics/article3628894.ece
Chris Green
York U.
Toronto, Canada
---
To make changes to your subscription contact:
Bill Southerly
([EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>)
--
Don Allen
Department of Psychology
Langara College
Vancouver, B.C., Canada
V5Y 2Z6
604-323-5871
---
To make changes to your subscription contact:
Bill Southerly ([EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>)
*********************************
Mark A. Casteel, Ph.D.
Associate Professor of Psychology
Penn State York
1031 Edgecomb Ave.
York, PA 17403
(717) 771-4028
*********************************
---
To make changes to your subscription contact:
Bill Southerly ([EMAIL PROTECTED]<mailto:[EMAIL PROTECTED]>)
--
Don Allen
Department of Psychology
Langara College
Vancouver, B.C., Canada
V5Y 2Z6
604-323-5871
---
To make changes to your subscription contact:
Bill Southerly ([EMAIL PROTECTED])
---
To make changes to your subscription contact:
Bill Southerly ([EMAIL PROTECTED])
