Page 100 - 4188
P. 100

98

         video  games—see  chapters  4  and  7).  This  was  taken  by  many  to  be
         “evidence” that media violence does not cause aggressive behavior.
                 Yet  many  causes  have  long-term  effects.  Consider  smoking  and
         lung cancer. Or consider water, salt, and your car. Over many years, cars

         that are repeatedly exposed to salt rust at a higher rate than those that are
         not exposed. But if you pour saltwater on your car, will you see it rust?
         No, it is a long-term effect. Some researchers have presented evidence

         that  the  effects  of  media  violence  may  be  long  term.  For  example,
         Centerwall  (1989)  has  documented  that  the  murder  rate  appears  to
         double  about  15  years  after  the  initial  introduction  of  television  to
         communities or countries. It has been hypothesized that about 15 years

         must elapse before the full effect is revealed, as that is the time it takes
         for a generation to grow up with the violent media and to reach a prime
         crime-committing  age.  If  this hypothesis  is  correct,  then  we  shouldn’t

         expect to see immediate effects. To the extent that we expect causation
         to appear as immediate or short-term effects, we may miss a number of
         important long-term effects.

                 Myth 7. Effects must be “big” to be important.
                 Many people have agreed that the accumulated research shows that
         there is a systematic effect of violent media on aggressive behavior, yet

         they also insist that it is not a large enough effect to be important. These
         discussions often include a statistical approach. For example, Ferguson
         (2002) notes that the amount of variance in violent behavior explained
         by  media  violence  in  meta-analyses  is  somewhere  between  1  and  10

         percent. This means that if we drew a circle representing all the reasons
         why  someone  might  act  violently,  media  violence  would  account  for
         between 1 and 10 percent of the pie. (It should be noted that some meta-

         analyses have reported larger numbers, and that there are a number of
         methodological reasons why these numbers may be underestimates; see
         Paik & Comstock, 1994, for a detailed explanation.) Ferguson (2002, p.
         447)  states  that  these  effect  sizes  are  “small  and  lack  practical

         significance.”  Ferguson  is  not  alone  in  making  this  type  of  value
         judgment  (e.g.,  Freedman,  2002),  but  it  is  unclear  on  what  basis  it  is
         made. In epidemiological terms, if only 1 percent of the people watching

         a  violent  TV  show  become  more  aggressive,  and  one  million  people
         watch  the  program,  then  10,000  people  were  made  more  aggressive.
         That does not seem to us to “lack practical significance.” Indeed, many

         (if  not  most)  medical  studies  on  the  effects  of  drugs  or  diet  are
         concerned  with  such  small  effects.  Supplementing  one’s  diet  with
   95   96   97   98   99   100   101   102   103   104   105