Report prepared for The Council of Europe

Directorate of Human Rights






 Guy Cumberbatch







B4 7ET


Tel/Fax:   0121 359 0844




Strasbourg: The Council of Europe 1995










Introduction and Overview                3


History of Media Violence                 4


Theories of Effect                               5


Criminal Statistics                               8


Natural Experiments                           10


Correlational Studies                          11


Longitudinal Studies                          13


Experiments on Children                    17


Experiments on Adolescents              20


Experimental Field Studies                 22


Delinquents                                         24


Content and Control                           27


Conclusions                                         30


Bibliography                                       35



Introduction and Overview


  Mass media violence is probably the most researched issue in mass communications.

  Few recent reviews even agree on the size of the research literature.  However, in

  1982 the National Institute of Mental Health listed well over a thousand research

  publications in this field.  Andison's (1977) conclusion was that three quarters (77%)

  of such studies claim that media violence is causally linked to real life aggression and

  this probably still holds true in most recent reviews  (eg Comstock, 1991; Harris, 1993;

  Huesmann and Eron, 1986; Huesmann et al, 1992; Huston et al, 1992; Gunter, 1994;

  Geen, 1994;  Lazar, 1994; Liebert and Sprafkin, 1988; Wartella, 1995).


  One simple summary of this research literature is offered by the meta-analyses of the

  statistical data provided in the various studies (for example Comstock, 1986; Hearold,

  1986; Paik and Comstock, 1994; Wood, Wong and Cachere, 1991).  These suggest

  that the overall strength of association between media violence and human

  aggressivity is in the order of between 4% and 10%.  In other words some 4-10% of

  the variability in human aggression can be predicted from exposure to mass media

  violence.  Thus the majority of authors conclude that media violence has a causal link

  with human aggression.


  However, sceptics have regularly challenged this interpretation, pointing to

  methodological problems and conceptual difficulties in drawing conclusions about the

  link between media violence and real-life aggression. Most critics suggest that at best

  the apparent association has nothing to do with a causal relationship between the

  media and real-world aggression (see Brody, 1977; Cashmore, 1994; Cumberbatch

  and Howitt, 1975; Duhs and Gunton, 1988; Freedman, 1984, 1986, 1988, 1992;

  Gadow and Sprafkin, 1989; Gauntlett, 1995;  McGuire 1986; Stipp and Milavsky, 1988;

  Wilson, 1994).  


  Fortunately, despite the daunting amount of published literature, the number of key

  studies is quite limited and the strengths and weaknesses of the various research can

  be well illustrated by reference to these.


  It is worth noting at the outset that research on mass media violence is predominantly

  North American, and is heavily influenced by psychological theory and methods.

  There is a notable absence of contributions from either criminology (eg Sampson and

  Laub, 1993) or from academics from within media studies (eg Buckingham, 1993).

  However, perhaps the most serious omission in all of the research literature is the

  absence of applied policy orientated work. Issues such as broadcaster/researcher

  dialogue that may develop strategies for dealing with media violence are almost









  History of Media Violence


  Concerns about media violence are far from new.  Indeed beliefs that popular culture

  may be to blame for society's ills can be traced back to St Augustine and Plato.

  However as Pearson (1983,1984) has observed such concerns have been well

  documented since the 16th century when "popular songs too often presented criminals

  as heroes" (Burke, 1978).  In 1776, Joseph Hanway suggested that debasing

  amusements in newspapers were among the causes of " . . . the host of thieves which

  some late years invaded us . . . ".  By the early 19th century, amusement houses were

  condemned for " . . . that early depravity and extent of juvenile delinquency that every

  magistrate acknowledges to exist . . ." (see Pearson, 1984).  In the 1840s "penny gaff"

  theatres were accused of encouraging immorality and imitative crime.  The Sixth

  Report of Inspectors of Prisons, dated 1841, pronounced " . . . if they do not directly

  corrupt the mind they tend to its vitiation by familiarising it with scenes of grossness,

  crime and blood, with a revolting coarseness . . . " (Worsley, 1849).  By the mid 19th

  century comic books, appearing as "penny dreadfuls", became a popular target for

  those anxious to explain an apparent rise in crime.  Greenwood (1869) reflected,

  " . . . a stray leaf of Panther Bill or Tyburn Tree may sow the seeds of immorality

  among as many boys as a town may produce . . . ".  In 1905 Charles Russell did not

  need to ask whether amusement halls and theatres caused crime in  Manchester,

  England.  He said " . . . horrible murders and terrible tragedies when acted before the

  footlights . . . " lead to " . . . so many instances of violence on the part of young men

  in the back streets of the city . . . ".


   Anxieties about the new medium of the moving picture appeared almost from the

  beginning of the cinema.  For example by 1912 the UK had set up the British Board

  of Film Censors in response to the criticisms of cinema films and the perceived need

  to protect people from their excesses.  The appearance of horror comics in the

  immediate post-war period quickly led to spirited campaigns against them by those

  concerned that they might corrupt the young.  In 1954, Frederick Wertham published

  a book entitled  The Seduction of the Innocent, arguing this case based on his

  experience as a psychiatrist.  However, even  as early as 1953 Klapper had attempted

  to summarise the various concerns about television as a social issue. He found that

  anxieties about crime and violence predominated as much as concerns about the

  amount of time children were spending with this new medium.  More generally there

  was the worry that children were being exposed to an adult world from which they

  really should be protected.  Finally, Klapper noted a recurrent theme against popular

  culture and a desire for "better" children's television, by which people usually meant

  they wanted more educational programmes and  "classics" (see Wartella, 1988).


  Over the decades of concern about the various media, social science research began

  to make its own distinctive contribution to debate.  Perhaps the most important

  development in this was the Payne Fund, set up in New York in 1928, which financed

  a rather well staffed series of twelve independent studies looking at the role of the

  cinema in the lives of young people.  A series of volumes was produced, together with

  a summary volume written by Charters in 1933.  The overall conclusions were

  somewhat muted, agreeing that there may be some effects produced by cumulative

  viewing, but regarding these as superficial changes such as in fashion rather than

  changes of any social or moral  significance.  The researchers as a group were very

  unwilling to attribute criminal delinquency to cinema going.


  Essentially similar conclusions were reached in Britain in 1951 by the Departmental

  Committee on Children and the Cinema (the Wheare Committee) which sponsored a

  large survey of all juvenile offenders who appeared before the courts during a six

  month period.  This produced an impressive sample of 38,000 young offenders under

  the age of sixteen. The committee concluded that in only 141 (0.4%) of these cases

  criminal behaviour might be related to cinema attendance (Home Office, 1951).  A

  parallel survey by the committee of 1,344 heads of children's clinics, officials from

  children's courts, teachers and other relevant specialists concurred that, by and large,

  it was difficult to lay any blame on the cinema for an increase in criminal or immoral

  behaviour. However these experts felt that there may indeed be superficial effects of

  the cinema, such as suggesting ways in which young people might commit anti-social

  activities, even if films did not encourage them to do so. 


  By 1960 the research literature on the effects of the mass media was beginning to

  grow quite considerably. For example in 1961 UNESCO compiled a bibliography of the

  effects of the cinema, containing around 300 references, two thirds of which had been

  published in European countries.  However, from the early 1960s onwards, the

  predominant research tradition became firmly established in the United States of




Theories of Effect


  Two psychologists are particularly associated with seminal work in this period:

  Albert Bandura at Stanford University  and Leonard Berkowitz at the University of

  Wisconsin.  Both researchers focused on film mediated aggression, but were initially

  more concerned with developing theoretical issues than in dealing with applied

  aspects of the mass media.  However, their laboratory based experimental approaches

  stimulated many other researchers to explore this promising new field.  Perhaps more

  importantly, the theoretical orientations of these two academics have helped shape the

  way in which later researchers have conceptualised the effects of the mass media.

  Bandura (1961) became interested in imitation and especially the way in which

  aggressive models may be copied by younger people.  In his experiments, beginning

  in 1963, he measured aggression using a large knock-down plastic clown, called a

  "bobo" or "bozo" doll, which when struck violently would bounce back up again due to

  a weighted base.  The design of the experiments was such that children would watch

  a film of a model addressing the bozo doll belligerently, " . . . the model pummels it on

  the head with a mallet, hurls it down, sits on it and punches it on the nose repeatedly,

  kicks it across the room, flings it in the air and bombards it with balls . . . " (Bandura,

  1973, page 72).  After being exposed to the model's curious antics, the children were

  then frustrated by being shown some attractive toys that they were invited to look at

  but not play with. The children were then led to a room containing many of the toys

  that the model had used and ostensibly left alone to play.  Hidden researchers

  observed the children in their free play and monitored their aggressive behaviour.

  Bandura found, as other researchers have since, that most children (up to 88%)

  readily imitate the aggressive models (Bandura, 1994).  Perhaps more impressively,

  even eight months later, young people may still retain 40% of the behaviours observed

  and are able to reproduce them when invited to do so (Hicks, 1968). 


  The research of Leonard Berkowitz was somewhat different, in that he used university

  students as the participants in his research.  Berkowitz was centrally concerned with

  developing learning theory as a perspective on aggression and his use of film material

  was almost accidental in being an extension of the techniques he had used to

  investigate the effects of aversive stimulation (Berkowitz, 1962).  Berkowitz suggested

  that aggressive thoughts could also be triggered by aggressive films, which " . . . can

  prime other semantically related thoughts, heightening the chances that viewers will

  have other aggressive ideas in this period . . . " (Berkowitz, 1984, page 411).  This

  priming of aggression of course is the opposite of catharsis, which argues that

  aggression is reduced by a vicarious viewing of aggression.  Indeed, the early

  research by Berkowitz was designed to test what he thought to be methodological

  weaknesses in previous demonstrations of a catharsis effect by Feshbach in the

  1950s and 60s.  Even today research evidence supporting the catharsis theory has

  been very sporadic notably by Grimm (1994) in Germany.


  One other early theory of media effects was that of desensitisation, or the blunting

  of emotional responses.  This seemed to be demonstrated by Lazarus in 1962, when

  he showed subjects a film of sub incision.  Essentially he found that if the audience

  had seen the film before they were less upset by it but this did not generalise to new

  material.  Amongst early work relevant to this issue must be mentioned that of

  Thomson (1959) who investigated the effects of crime dramas on Australian school

  children.  He reported that those who watched a lot of television and films seemed to

  be much less emotionally responsive to thriller films than those who watched little. He

  concluded that this phenomenon might be connected with their increased experience

  reducing their sensitivity to such material.  However this view has not been supported

  in other surveys (Cantor, 1994).


  In the intervening years there are perhaps only two theoretical developments worthy

  of note. The first, by George Gerbner (1972, 1990), is a model that attempts to

  describe the impact of television on society by a close analysis of television content.

  He suggested that the main influence of television lies in its ability to convey ideas

  about social behaviour, social norms and social structures.  The more people watch

  television, the more their social and cultural beliefs will be changed.  At its simplest

  level, for example, the high frequency of violence on television should provide a

  message to people "the world is a violent place" and encourages a fear of crime.

  Gerbner focuses on what he calls the "symbolic messages of television", which are

  revealed by coding the content of television drama ("message system analysis").  The

  process whereby public attitudes and beliefs may be shaped by television is referred

  to by Gerbner as "inculturation" (see Gerbner, 1990).


  In 1986 Huesmann proposed a theory of cognitive scripting. He suggested that

  observing television violence allows viewers to learn complex behavioural scripts

  about how to act violently.  People may absorb these scripts more readily if they

  identify with the aggressive television character and if they feel that the portrayal is

  realistic.  Whether they use the script or not will depend on the amount of similarity

  between the situation at the time of retrieval and the situation at the time the script was

  encoded in memory.  This approach heavily draws on theoretical developments within

  the psychology of memory and language (for example Abelson, 1976).


  The theories of Gerbner and Huesmann are far more sophisticated than the early

  notions of media effects which surround most empirical research in this field.  The

  simplest, and perhaps most obvious, test of the theory that television causes real

  world aggression is to examine criminal statistics before and after the introduction of

  television.  Do crime rates rise more rapidly when television penetrates a culture?

  Criminal Statistics


  Here the evidence against television would at first sight seem damning. Centerwall

  (1989) claims that in virtually every country following the introduction of television,

  violent crime has shown substantial and sustained increases (Centerwall 1989).  Of

  course any society embracing television will simultaneously be undergoing other quite

  profound technological and social changes, especially since most Western countries

  introduced television soon after the Second World War.  Indeed, as Winston (1986)

  has argued in his book Misunderstanding Media, the Second World War and its

  occupational capacity for producing cathode ray tubes was probably central to the

  rapid development of television in the post-war period.  Moreover Centerwall provides

  data on only three countries to support his case.  More information is desirable before

  we can conclude that television may encourage violent crime.


  Fortunately there have been a number of studies that help illuminate the problem.  All

  three were carried out in the USA and thus may be of limited relevance to other

  countries.  Nonetheless, they remain interesting.  The first study, by Hennigan et al

  (1982), used archival data to look at the association between crime statistics and the

  introduction of television.  It was convenient that in the United States the Federal

  Communications Commission operated a freeze on television station licensing in the

  late 1940s.  This allowed the researchers to carry out an interrupted time series

  analysis, switching replications in which the roles of the treatment group receiving

  television and the control group that did not, reversed over time. 


  Hennigan et al amassed statistics for crimes in four categories: crimes involving

  violence; larceny (theft); auto theft and burglary.  The results of this detailed analysis

  are quite interesting.  First of all there were no significant shifts at all for violent crime

  or burglary, but larceny (theft), showed quite a significant increase while auto theft

  produced only minor changes.  Although modest in the time period covered, this study

  therefore does not support the main point raised by Centerwall, that violent crime has

  increased following the introduction of television. 


  The second study, by Clark and Blankenberg (1971), looked at the relationship

  between crime statistics and the violent content of television.  They used the TV

  programme magazine, TV Guide, to assess the violent content of television for each

  of the years between 1953 and 1969.  As with many such magazines, TV Guide offers

  plot synopses and details of the programmes thus providing a rudimentary means to

  judge the violent content of television. Programmes could be classified in terms of

  whether they were likely to be violent or not.  Other research by the authors

  demonstrates that this leads to an underestimation of the violence in programmes, but

  nevertheless it is usually reasonably well correlated with the total amount of violence

  on television in any year. 


  Clark and Blankenberg found that there was a considerable fluctuation in the amount

  of violence shown over the years.  Most interestingly, they reveal that the violent

  content of television tends to be cyclical; it goes up and down and reaches peaks

  roughly every four years.  For researchers the value of such variation is that it can

  then be correlated with the annual crime statistics, in order to check whether crime

  statistics show similar cycles.  No relationship at all was found between the violent

  content of a year's television programming and the crime figures of that year.

  Moreover, in carrying out lagged analysis between one year's television violent

  content and the crime statistics of the following year, there was similarly no



  The final study, by Messner (1986), utilised the geographic variations in data across

  the United States.  Because the rate of crime in some states is considerably higher

  than in others, hypotheses can be tested about factors which might "explain" these

  variations (cf Baron and Strauss, 1989). Similarly exposure to television may vary in

  a similar fashion. Messner used the data from the 281 Standard Metropolitan

  Statistical Areas (SMSAs) to examine the television viewing frequency as given by

  Nielsen data in order to estimate the relative exposure to violent programmes across

  those areas.  In addition he looked at the rates of four types of crime: homicide

  (murder); rape; robbery and aggravated assault across these metropolitan areas.

  Messner's conclusions were surprising: fewer crimes occurred in areas where

  television viewing was greater than average, and more occurred where it was lower

  than average.  Messner thus concluded that there is no relationship between television

  violence viewing and delinquent behaviour. He attempted to explain this by suggesting

  that young people are likely to be socialised to act criminally by association with their

  peer group, rather than spending time at home watching television.  This is an

  intriguing suggestion that will be returned to since it is supported by more recent

  research on juvenile offenders.


   Although conceptually quite different from the above studies, it is worth noting here

  the quite considerable literature generated by Phillips on the contagion of media

  violence.  This research uses archive data to test the hypothesis that mass mediated

  events, such as prize fights, lead to an increase in homicide.  For example, he looked

  at the fluctuations in daily homicides in the United States before and after every

  heavy-weight championship fight between 1973 and 1978.  After quite complicated

  analysis of the data, he reports a small but significant "effect" of prize fights on the

  homicide rate.  Each prize fight led to approximately 12 more homicides than

  expected.  Phillips claims that a vast range of media events produces similar imitation,

  including publicised suicides, public executions, murder stories and prize fights. In

  each of these areas Phillips has produced an impressive number of supportive

  publications (eg Lesyna and Phillips, 1989; Phillips, 1982, 1983, 1986;  Phillips and

  Hensley, 1984). 


  While this kind of research evidence lends support to the theoretical position of

  Bandura on imitation, there have been a considerable number of contradictory results

  (eg Baron and Reiss, 1985; Berman, 1988; Kessler et al, 1988; Kessler and Stipp,

  1984; Messner, 1986; Phillips and Carstansen, 1986; Phillips and Paight, 1987; Platt,

  1987; Schmidtke and Haffner, 1988; Stack, 1992). 


  Given the enormous interest this kind of research has attracted and the controversy

  it has created, it would be misleading to offer a simple summary.  Much of the

  controversy surrounds the statistical methodology.  For example Baron and Reiss

  (1985) replicated Phillips study on prize fights, but invented the date of the prize fight

  to see whether the non existent fights were followed by increases in the levels of

  homicide.  They found evidence that this was the case, which means that the non-existent prize fights had exactly the same "effect" as the real ones.  There is no doubt

  that in the face of such and similar criticisms Phillips has always provided a spirited

  defence of his research position.


  However, we should remember that the "mass mediated" events with which Phillips

  deals are not necessarily only mediated by television, they may include newspapers

  or observers at those events or even hearsay evidence about those events. Moreover

  claims that anti-social events reported in the media produce "copy cat" effects ignore

  the role of news values in this.  A wave of media reports of dangerous dogs savaging

  children may be due to the newsworthiness of such stories following the first report

  than on imitative behaviour in the canine species.  In this sense the research does not

  tell us very much about the role of the mass media in society, still less about the role

  of television in creating anti-social behaviour. 



  Natural Experiments


  The introduction of television could have allowed some sophisticated research to

  examine the impact of television on society.  However it is a matter of some regret that

  so few studies exist.  Happily, the unique opportunity to carry out before and after

  studies was not missed either in North America or in Britain.  Large scale and fairly

  comprehensive evaluations were conducted by Himmelweit, Oppenheim and Vince

  (1958) in Britain, and by Schramm, Lyle and Parker (1961) in Canada and the United

  States.  These are both important studies and make fascinating reading, although

  unfortunately the measures of aggression taken were fairly perfunctory in both studies.

  Those taken by Himmelweit involved only one item on a questionnaire given to

  teachers, who were asked to complete check lists on the children.  However, no effect

  of television was found on children's aggression.

  Schramm, Lyle and Parker's measures of aggression were more sophisticated  the

  findings a little different and the conclusions possible seem contradictory.  In their first

  study, young children between the ages of eleven and twelve, in what was called

  Radio Town, were more aggressive than those in Teli Town, while no differences

  existed between the groups of older children.  In their second study, a weak trend was

  found for heavy viewers of television to be more aggressive than light viewers.


  Opportunities to evaluate the introduction of television have become increasingly rare

  and precious few other studies exist.  The most recent publication, (Joy, Kimball and

  Zabrack, 1986) reports the results of a natural experiment taking place well over a

  decade earlier in Canada.  Here a small logging and farming community hidden in a

  valley and shielded from the transmitters, was suddenly exposed to television for the

  first time.  The before and after measures taken suggest an increase in both physical

  and verbal aggression compared with two control communities, one that had only one

  television station (Unitel), and another that had more than one (Multitel). 


  While this in some ways seems the best evidence yet reviewed of an increase in

  aggression occasioned by television, the authors interestingly interpret their results

  not in terms of the content of television but more in terms of the general presence of

  television, leading to an increase in material values etc.  As Murdock (1988) has

  observed, it seems such a shame that this study did not allow a more substantial

  analysis of the community, which had a population of just 658 people. Notel was small

  enough to have made the construction of a multilayered account of the community in

  the process of change a real possibility.

  The last study examining the introduction of television was carried out by Granzberg

  and Steinbing (1980) in a Cree Indian community, which they compared with a control

  Indian community and a control Euro-Canadian community. No pre or post differences

  were found in the levels of aggression between the experimental and control

  communities.  Thus in this case television was not associated with an increase in

  aggressive behaviour.  Nonetheless, Granzberg and Steinbing found that when they

  looked at the correlation between daily exposure to television and aggressive attitudes

  a significant relationship emerged.  Unfortunately this finding is only one of correlation

  rather than change and is a subject that will be examined under the next section.



  Correlational Studies


  One of the earliest surveys to focus on the possible relationship between television

  and delinquency was carried out in Germany by Maletzke (1959).  He interviewed two

  matched groups of 200 youths, between the ages of fifteen and twenty. They were

  divided according to whether or not they were regular television viewers. He found no

  evidence of any different delinquent propensity between these two groups, or indeed

  anything that suggested that television had any clear effect at all on the young people.


  Around the same time in the United States Pfuhl (1960, 1970) undertook a more

  elaborate piece of research, examining a whole variety of media consumption by

  young people between the ages of fifteen and eighteen.  He assured the participants

  of anonymity, since some of the measures referred to delinquent behaviour, such as

  purposely damaging or destroying public or private property.   There was no variation

  in delinquency according to a wide variety of measures of television viewing, including

  time spent watching television, favourite types of television programmes and exposure

  to television crime dramas.  Thus Pfuhl was quite adamant that television was not

  implicated in delinquent behaviour. 


  In 1963 Eron reported the results of a study of 875 young people, aged between eight

  and nine. He had measured their aggressiveness using peer ratings (that is the young

  people nominated those who were more aggressive at school) and measured their

  television exposure by reports from the children's mothers.  First of all he found there

  was a negative correlation between the amount of time children spent watching

  television and peer rated aggressiveness: the more they watched the less aggressive

  they were. 


  On the other hand, Eron did report that those boys who preferred aggressive TV

  programmes, as reported by their mothers, were more aggressive at school, as

  reported by their peers. On the whole these early studies gave only limited support for

  concerns about television. 


  However the reassuring pattern of research findings was challenged in 1970s when

  a spate of studies was published, ostensibly demonstrating associations between

  television violence viewing and aggressiveness. For example, McLeod, Atkin and

  Chaffee (1972) in a survey of 600 adolescents in Maryland and in Wisconsin, found

  that exposure to violent television was positively correlated with self-reported

  aggression.  Among the various sets of data produced the strongest association

  between television and aggression was in young girls.  However, McLeod, Atkin and

  Chaffee report a number of other findings, for example, that total television viewing,

  violence viewing and self-reported aggression all declined from junior to senior high

  school and that boys watched considerably more violent television than girls.  It may

  be that the relationship between television and aggression, as reported by the

  researchers, is an artefact due to not making adequate allowance for the sex of the

  child, and their age.  For example, when the various groups, formed by sub-dividing

  the sample by sex and by school placement, are examined only one correlation out of

  the eight reported results reaches statistical significance.  One other way of viewing

  this data is to examine the total number of correlations reported between the various

  measures of aggression and television viewing.  Of the twenty described, nearly half

  were significant, but unfortunately only one of these is significant in both the Maryland

  and the Wisconsin samples. This would seem to suggest that the reliability of the data

  does not reach the standards that we would need for policy recommendations.


  In the same year McIntyre and Teevan (1972) reported an even larger study of 2270

  adolescents, again in the Maryland area.  Their research measured a variety of

  television viewing habits and various delinquent activities and concluded that

  preference for violent television programmes was related to serious aggressive

  delinquent acts, but not to petty delinquency.  Not surprisingly, McIntyre and  Teevan

  found that boys were more delinquent than girls, while black and lower socio-economic

  groups contributed most to the overall deviancy.  As with the previous study, when

  age, sex and race are controlled, the reported correlations between television

  exposure and delinquency become considerably weaker and statistically insignificant,

  with only one exception.  This would again indicate that the results do not achieve the

  level of reliability desired. 

  Somewhat similar results were reported by Robinson and Bachmann in a survey of

  1500 adolescents. They again found a small correlation between violent programme

  preference and aggressive attitudes, but no significant results due to overall television



  These findings were again echoed in the latest study by Thornton and Voight (1984)

  who drew a probability sample of 3500 males and females.  Their measures of

  television exposure were fairly cursory, including the amount of television viewed and

  the young people's four favourite programmes.  However, the delinquency measures

  were more extensive, covering 27 items.  The associations reported were not with total

  television viewing, but only with preferences for violent television programmes.  The

  associations were generally very weak, but stronger with serious delinquent

  behaviour, such as criminal damage, than with more petty delinquency.


  The absence of a relationship in these studies between total television viewing and

  aggressive delinquent behaviour is puzzling since the more people watch television,

  the more likely it is they will watch television violence.  It could be, therefore, that the

  programme preference data merely indicate that aggressive people are attracted to

  and enjoy aggressive television programmes. 


  Although various studies lend support to this idea, perhaps the most elegant test of

  this hypothesis is provided by Lynn, Hampson and Agahi (1989) in a large study

  carried out in Northern Ireland.  They surveyed over 2000 children and identified

  siblings within that sample to focus the research on differences between brother and

  sister pairs in terms of their relationship to television and their aggressivity.  The

  results were quite striking:  there was no evidence at all  that the amount of viewing

  television violence had any effect on aggression. However, enjoyment of television

  violence was correlated with aggression.  Overall the best predictor of aggressivity in

  this sample was psychoticism.  Thus Lynn's data provides strong support for the

  hypothesis that aggressive children tend to enjoy aggressive television.  Lynn's

  conclusions were circumspect:  he suggests that enjoying television violence may

  have an influence in increasing aggression in young people, but the evidence for this

  is weak and remains speculative.


  It seems clear that the correlational studies have not helped much to illuminate the

  processes whereby television violence might be related to aggressive behaviour.

  There is some doubt as to whether television viewing in itself is the cause of the

  relationship between the enjoyment of television violence and aggressive behaviour.

  Perhaps the best way to examine the dynamics of such a relationship is in a

  longitudinal study, following children through over time.  We turn now to these.



  Longitudinal Studies


  In 1972 Eron, Huesmann, Leftkowitz and Walter reported the results of a longitudinal

  study that had followed up on Eron's original 1963 sample.  Eron had measured

  children's aggression and television violence exposure at the age of eight or nine and

  surveyed them again ten years later.  Essentially they claimed there was a delayed

  action effect, where early preferences for television violence correlated with

  aggression at the age of eighteen or nineteen.  This is true possibly for one of the

  measures of aggression for boys, but two of the other measures of aggression did not

  show a relationship and none of the measures showed this effect for girls. 


  The research continued and in 1982 Huesmann reported the results of a later follow-up study that failed to find any effects for either sex.  Indeed, at adulthood the

  correlation between television and aggression was slightly negative, but he did not

  attach any importance to this. 


  It is perhaps unfortunate that the measures used in the Eron study are not more

  comprehensive.  Information on early television violence viewing was obtained from

  parents and, as Kay (1972) has pointed out, they are not likely to provide very valid

  data.  Parents might, for example, guess that aggressive children preferred aggressive

  programmes; indeed, the possibility that aggressive personality explains early

  television violence viewing and later aggression is an issue which is not taken

  seriously in Huesmann's report.


  Singer and Singer carried out two separate year long studies of pre-schoolers

  investigating the relationship between television viewing and their aggressive

  behaviour (Singer and Singer, 1980;  Singer and Singer, 1981).  The mothers of  the

  children kept daily logs on the children's behaviour for a two week period several times

  during the year.  Additionally trained observers recorded details of  the children's

  behavioural and effective responses.  Their conclusion was that heavier viewing of

  television significantly predicted later aggression.  However in a detailed analysis of

  the results they became convinced that total television may be less important than

  what children watch.  For example children who were highly aggressive, including

  those who were very low television viewers, average four times as much time watching

  adult orientated action adventure programmes as did low aggressive children.  The

  lower aggressive children tended to watch programmes like Mr Roger's

  Neighbourhood significantly more often than the highly aggressive children.  However

  they also noted that later aggressive behaviour that was predicted by " . . . early heavy

  viewing of public television's fast-paced  Sesame St as well as more violent cartoons

  such as Super Heroes and Woody Woodpecker . . . " (Singer and Singer, 1986, page



  In Germany a longitudinal study of 2500 eleven to fifteen year old children was carried

  out over a period of two years in three waves (Groebel, 1981;  Krebs, 1981).  The

  measures of aggression were classified as reactive (instrumental aggression) or

  destructive (spontaneous aggression).  Television viewing was related to reactive

  aggression but not to destructive aggression.  However additional analyses reveal

  some other interesting findings.  Boys with low self-esteem and high social fear

  frequently had higher TV viewing scores than other groups.  The researchers interpret

  this as an escapist function of television.  Probably the largest longitudinal study is

  that carried out by Milavsky (Milavsky, Stipp, Kessler and Rubens, 1982).  This took

  more than ten years to complete and involved surveying 3200 young people over a

  three year period.  Data was obtained at six points in time through questionnaires

  administered to the children in their classrooms.  Parents and teachers were also

  interviewed so that any effects could be analysed in the context of both the children's

  homes and the school environments.  As Milavsky pointed out (Stipp and Milavsky,

  1988) " . . . one of the reasons that the study took so long to complete was the results

  were weak and inconsistent and therefore difficult to interpret . . ." (page 164).  They

  concluded any effect of watching television violence on children's aggression either

  did not exist or if it did it was so tiny it was hardly measurable.  Opinion quickly divided

  in the academic community over whether Milavsky's findings did show some effect

  after all or whether they were confirmation that no real link existed between television

  violence and aggressive behaviour (Cook, Kendzierski and Thomas, 1983;  Felson

  and Morgan, 1984;  Kenny, 1984;  Rossi and Wright, 1983). 


  The most detailed research endeavour in this field is that by Belson (1978) who

  studied 1565 boys aged thirteen to sixteen in London.  Although not a longitudinal

  study Belson asked children to recall programmes that they had seen in their youth

  and used this data to attempt to predict their current aggression.  Given the enormous

  amount of data collected on the boys this must have involved some of the lengthiest

  interviews carried out in social science.  Belson tested twenty two hypotheses

  concerning the possible relationship between television violence and children's

  aggression.  The aggression was measured using 45 different types of aggressive

  behaviour.  These included "I've had an argument with somebody", "I have been

  violent in self-defence", "I've shot and airgun or catapult or arrow at someone or

  something" and "I've cut someone with a knife or a razor or a glass".  They were asked

  if they had done it in the last six months and then how often they had done it in the last

  six months. 


  The most important element in Belson's research was the way in which he attempted

  to match heavy and light viewers to television violence using a sophisticated system

  which controlled for over 200 different variables such as social class, education in

  order to overcome the serious problem that any correlation between delinquency and

  exposure to television violence could be due to a third variable.  It would be churlish

  to criticise Belson for not being thorough enough in this search for third variables

  though his matching list is far from exhaustive.  Belson's conclusions were confident

  and dramatic.  Boys with high levels of exposure to television violence committed forty-nine per cent more acts of serious violence than those who saw little television

  violence.  Belson went on to list policy recommendations and suggest that violence

  should be reduced on television, especially in " . . . plays or films in which violence

  occurs in the context of close personal relationships . . . ", " . . . programmes

  presenting fictional violence of a realistic kind . . . " and so on (eg page 520).  This

  reads as an impressive list of specific recommendations which people may desire to

  pinpoint the controls on television. 


  Unfortunately closer examination of the vast data that Belson produced urges more

  caution.  For example the graphs for the full sample (pages 380-382) show that the

  results are far from as simple as his conclusions imply.  In these graphs where

  exposure to television violence is plotted against violent behaviour it is clear that the

  relationship is curvilinear.  Thus very low viewers of television violence are slightly

  more aggressive than moderate viewers.  More importantly very high viewers of

  television violence are less aggressive than the moderate to high exposure group, fifty

  per cent lower in fact.  Moreover in Belson's data exposure to non-violent television

  is also linked to aggressive behaviour as indeed are comics and comic books and

  even newspaper readership. 


  Another serious problem with Belson's research relates to the measures taken of

  television violence exposure.  First of all the list of programmes presented to children

  included some which had ceased to be broadcast when the children were only three

  years old (Murdock and McCron, 1979).  Under these circumstances the validity of the

  responses may be called into doubt.  Belson attempts to reassure critics on this point

  by demonstrating that the responses were "reliable" (ie that on a second occasion

  boys produced the same responses as on the first one).  Although this is a

  conventional measure of reliability in research methods it does not ensure validity. 


  For example children might repeatedly assert they are Martians which would on this

  criteria have to be considered as a reliable response, but we might assume that the

  response lacked some validity. There are of course various ways around this problem.

  One solution adopted by Milavsky (Milavsky, Kessler, Stipp and Rubens, 1982) in a

  very similar study to Belson's is to present children with some non-existent television

  programme titles.  If children claim to have seen these then they may be considered

  to be providing invalid responses.  What is interesting about Milavsky's research is

  that he initially found a weak connection between television violence exposure and

  aggression but when he removed those respondents who gave invalid responses the

  correlation was dramatically reduced.  In other words the link found by Belson and

  others may well be due to the inadequacy of the measures of television violence

  exposure taken. 


  As a final point we should note that despite Belson's potential database of 1565

  respondents the actual numbers contributing the claimed link between television and

  serious acts of aggression must be considerably smaller.  After all few children

  committed very serious acts of aggression.  It is not clear from Belson's data exactly

  what the sizes of the various sub samples were but his cut-off point for matching was

  only 30 cases. 


  The most cited of longitudinal studies are those associated with Huesmann and Eron

  (1986) who orchestrated cross-national comparisons in the United States of America,

  Australia, Finland, Poland, Israel and Holland. Although the designs vary slightly

  between countries they are essentially similar in that peer and self-ratings were used

  in aggression and self-ratings of television exposure.  In the USA for example 34 lists

  were presented to the children who were invited to select from each list one or two

  programmes which they watched and then to say how often they watched them.  In

  their account of this research the authors repeatedly assert that television violence

  viewing predicts later aggression in the children.  However the pattern of findings is



  The Dutch researchers considered their results did not show any effect of television

  violence and refused to publish within Huesmann and Eron's book and report their

  results elsewhere (Wiegman, Kuttschreuter and Barda, 1992).


  In Australia there were no significant correlations between early television violence

  viewing and later aggression. 

  In the USA after controlling for initial aggression the relationship between television

  violence and viewing was significant only in girls. 


  In Israel significant correlations were found in the city sample but not in a sample

  drawn from a kibbutz. 


  In Poland while there were correlations where " . . . a greater preference for violence

  viewing was predictive of greater aggression . . . ".  The author Fraczek adds

  " . . . nevertheless the effects are not large and must be treated cautiously . . . ".


  The Finnish researchers Lagerspetz and Viemero conclude " . . . our study in Finland

  can be taken to corroborate the previously obtained results that the amount of

  aggressive behaviour in children is related to their viewing of violence on TV . . . ".

  However these results are not significant for girls and the correlation quoted for boys

  depends on a new variable computed from the product of television violence viewing

  and identification with television characters.  It is apparent from the full report, which

  they published independently (Viemero, 1986), that the correlation of violence viewing

  with later aggression is actually negative (-.324).  In other words the more children

  watch violent television the less aggressive they were later.  Moreover the measure

  of identification clearly indicates that for boys it was identifying with female characters

  that correlated with aggressive behaviour (+.715).  Under these circumstances it is

  difficult to concur with the author's conclusions. 


  In other respects the details reported of the analyses carried out are inadequate to

  reach firm conclusions about the role of third variables.  For example in the Polish

  study Fraczek notes that academic achievement halved the correlation between

  violence viewing and later aggression but academic achievement does not seem to

  be have been controlled for in the final regression analysis. 


  Thus overall the longitudinal studies have repeatedly pointed to some weak

  associations between television and aggression but gloss over some serious

  contradictions in the data and do not aid in the understanding of what this relationship

  is nor indeed can they be taken as the basis for policy recommendations. Links

  between media violence and aggression appear as a "Will o' the Wisp" phenomena

  never in the same place at the same time.  As many authors have suggested (eg Prell,

  1992) young people may be more impressionable than older age groups and so

  longtitudinal studies would seem the ideal way to follow through possible media effects

  but these remain to be convincingly and reliably demonstrated.



  Experiments on Children


  Following the pioneering laboratory experiments on young children by Bandura in the

  early 1960s a plethora of studies followed often utilising the convenient university

  creche facilities enjoyed by academic psychologists.  The willingness of most young

  children from the age of three to imitate models in a laboratory quickly led the research

  to moving on from asking whether children can and do imitate to asking questions

  about the circumstances which optimise such imitation.  Taken as given that children

  will imitate models and film models in a laboratory, the question became will the

  children imitate more if the model is rewarded, rather than punished, will the children

  imitate more if the model seems to be of a higher status and successful, will the

  children imitate more if the model seems to be similar to the child?  And so on.  It is

  unfair to summarise all of this research by a simple statement, but the evidence from

  all of the research is that children imitated the models almost regardless of the

  conditions (Bandura, 1986, 1994; McHan, 1985).


  However, over the years, there has been a growing dissatisfaction with the ecological

  validity of the laboratory experiment. While there is no doubt that children can learn

  by observation, and this must be part and parcel of the way in which children are

  socialised, children do not typically imitate all that they see.  There is a discrepancy

  between the very high incidence of imitation in the Bandura style experiment and what

  children typically do after watching television (play normally, have their tea and so on).

  Undoubtedly children may imitate some things they see but the novelty of the 'bobo'

  doll was a crucial factor in Bandura's experiments. For example Kniveton and

  Stephenson (1970) found that children who were not familiar with the doll imitated five

  times more than children who had previous exposure to it.  Indeed, Nobel (1975)

  commented, " . . . in my own studies where children watch media violence in small

  groups I have rarely found more than 5% imitation after viewing . . . " (Nobel, 1975,

  page 134).  Nobel suggests that even young children taking part in the laboratory

  experiment understand that they are expected to play a particular role.  For example

  he quotes one shrewd four year old who on first arriving at the laboratory for a

  modelling experiment was heard to observe to her mother, "Look Mummy, there's the

  doll we have to hit!" (Nobel, 1975, page 134).


  Despite Nobel's preference for carrying out research in more naturalistic settings it is

  still the case that field experiments with young children have found associations

  between exposure to aggressive material and children's anti-social behaviour: more

  aggression has been reliably found following exposure to television violence (eg

  Sanson and Di Muccio, 1993).  However, Gadow and Sprafkin (1989, 1993) point out

  a peculiar problem that seems to have escaped the notice of so many reviewers.  In

  the 20 field experiments available, which they carefully reviewed on the short term

  effects of television on children's behaviour, the stimulus materials seem to show quite

  complicated and controversial results.  By and large the finding is that while

  aggressive film content often produces elevated levels of anti-social behaviour, the

  control (non-aggressive) material produces sometimes even greater amounts of

  aggressive behaviour.  In other words, the field experiments carried out with young

  people do not implicate television violence per se in the anti-social aggressive

  behaviour demonstrated in free-play.  The authors point out that even television

  programmes which are specifically produced with the aim of encouraging pro-social

  behaviours in children, may encourage aggressive behaviour.


  In this context it is worth remembering the experiment by Coates et al (1976) which

  found that viewing Sesame St and Mr Roger's Neighbourhood resulted in an almost

  threefold increase in aggression in their more aggressive pre-school children samples.

  Singer and Singer in 1986 reached a somewhat similar conclusion, suggesting that

  " . . . later aggressive behaviour was predicted by earlier viewing of public television's

  fast paced Sesame St, as well as more violent cartoons, super-heroes and Woody

  Woodpecker (Singer and Singer, 1986, page 113).


  As Gadow and Sprafkin argue, the policy implications of such research evidence are

  troublesome.  Interestingly, while they conclude that the findings from field

  experiments with young people offer little support for the media-aggression

  hypothesis, they also suggest that aggression laden programming can be justifiably

  rejected on other grounds, such as aesthetic, humanistic and philosophical.  Of course

  in reviewing field experiments Gadow and Sprafkin were concerned only with the

  short-term effects of television and films. There is a more general issue as to whether

  children may identify with aggressive television characters and by so doing change

  their attitudes and values in such a way that makes aggression more likely in the long

  term.  Whether or not children immediately imitate aggressive characters in play may

  be less important than whether they identify with such aggressive television characters

  and, as Huesmann has argued, learn aggressive cognitive scripts that encourage

  them at some later date to reproduce that aggressive behaviour. 


  Unfortunately, Bandura in his various theoretical offerings (eg Bandura, 1986) still

  regards identification and imitation as more or less synonymous.  In Bandura's

  formulation imitation is less a process of media effect than an effect in itself as

  measured by matching responses of the observer to the model.  However,

  identification where an observer desires to become like the model could be a process

  of effect (Barker, 1989).  To date little research attention has been paid to this issue

  despite frequent references to identification in the writings of Huesmann (eg

  Huesmann and Malamuth, 1986).


  The most focused experimental research in this area examined the potential influence

  that identification with an aggressive film character might have on children's attitudes

  to and moral judgements about aggression (Howitt and Cumberbatch, 1972;

  Cumberbatch and Howitt, 1974, and Howitt and Cumberbatch, 1975).  Children were

  shown a lengthy film clip to elicit identification responses.  Films used included Chick's

  Day, Billy Budd and The Virginian. Identification was measured by a series of

  questions asking, for example, "How much do you like Chick? How much would you

  like Chick as a brother? How much would you like Chick as a friend?" and so on.  In

  order to test whether identification with a film character would lead children to

  becoming more tolerant of that character's aggressive behaviour, two versions of a

  questionnaire were prepared.  Each described an extended aggressive scene which

  began pro-socially and ended up with excessive force being used, where a man

  becomes knocked unconscious.  One version attributed the aggressive behaviour to

  "Chick" but the control group were presented with the violent scene attributed to a

  " . . . John Davies who was not in the excerpt shown . . . ".  The crucial question was

  whether children who identified with the film character would become more tolerant of

  that character's aggressive behaviour compared with when the same behaviour was

  attributed to someone with whom they had not identified.  In none of the experiments

  was there a significant difference between the experimental group and the control

  group.  In other words, identifying with the television character did not lead to children

  becoming more tolerant of that character's aggressive behaviour. 


  However, each of the experiments demonstrated that children who identified with

  aggressive characters would tend to be more tolerant of aggressive acts, regardless

  of whether they were attributed to the character with whom they had identified, or to

  a control character whom they had not encountered before.  In other words, it would

  seem that children who are tolerant of aggression may more readily identify with

  aggressive television characters.  However crucially, identifying with aggressive

  television characters does not cause them to become more tolerant of aggressive

  behaviour.  This finding is reminiscent of that offered  by Lynn et al (1989) who

  demonstrated that television violence appealed to those of aggressive disposition but

  was unable to show evidence of the media aggravating anti-social dispositions.


  This is not to suggest that research evidence is inconsistent with the thesis that media

  violence causes aggression but rather that the evidence is open to other



  The range of methodologies adopted to explore media effects has been somewhat

  constrained but the research by Molitor and Hirsch (1994) is worth mentioning on

  children's tolerance of 'real-life' aggression following exposure to media violence.  This

  study successfully replicated an earlier one by Drabman and Thomas which required

  young children to monitor the behaviour of other children.  Children participating in the

  research were asked to do this using a closed circuit television system and to call for

  help if the behaviour got out of hand.  The effects of previously viewing aggressive

  media fare was to delay children's calls for help.  This is consistent with desensitation.

  However experiments of this kind are very susceptible to how the participants interpret

  the task demands (Nobel, 1975).  Thus if children are shown aggressive films by an

  experimenter they may well interpret this as reflecting a relatively tolerant attitude to

  aggression (cf Borden, 1975).


  All in all while short term effects are easy to demonstrate these appear to dissipate

  quickly (Frydman, 1995) and offer little support for concerns that television violence

  has a long-term influence on the aggressivity of individuals still less on the social

  problem of violence in society.



  Experiments on Adolescents


  Berkowitz has been a leading researcher within the field of aggression for more than

  three decades (Berkowitz, 1962-1993).  From the outset his theoretical formulations

  involved the use of quite complex designs, since he argued that media aggression

  would increase or prime aggressive drives only if it were justified media aggression

  and if the person watching it was already aggressive. 


  The general procedure adopted by Berkowitz involved six groups of participants,

  (university students), who first of all received an intelligence test.  Half of the groups

  were angered by an experimenter who insulted the intellectual competence of the

  subjects, while the other group were treated in a neutral fashion.  Following this the

  groups were again split.  Half of the group saw a control film about canal boats in

  England while the other half saw a seven minute film clip of a prize fight scene taken

  from Champion which stars Kirk Douglas.  One experimental group was informed that

  the aggression against Kirk Douglas in the film was justified  (in order to reduce the

  participants' aggression anxiety) while the other group were told that it was not (in

  order to increase their aggression anxiety).  At the end of the experiment the

  participants were invited to complete a questionnaire on what they thought of the

  experimenter. They were told that this would be sent to the head of the department.


  The results from this research indicated that those angered by the experimenter made

  more adverse comments about him than those treated in a neutral way, not

  surprisingly.  In the angered group those who had been instructed that the aggression

  in the film was justified made significantly more aggressive responses than either the

  control film group, or the unjustified aggression film group.  The films did not have any

  differential effect on the non-angered groups. Thus Berkowitz concluded that

  " . . . media aggression depicted as being justified has the greatest probability of

  leading to aggression when the audience is already angry  . . . " (Berkowitz, 1962,

  page 243). 


  Later experiments became more complex and indeed more contrived. Thus Berkowitz

  and Geen (1967) hypothesised that when given the opportunity to aggress, subjects

  would do so more strongly if the victims were linked with the aggressive film in some

  way.  This experiment involved the second experimenter being introduced as either

  'Kirk' who was boxer or 'Bob' who was not.  After seeing the Kirk Douglas film

  sequence, where Kirk plays the role of a boxer, subjects by a happy coincidence were

  allowed to give electric shocks to the experimenter in the learning task.  Subjects who

  received the name mediated treatment produced the most aggressive responses.

  Similar findings were reported by Berkowitz and Geen (1966). 


  Although Berkowitz is clear about the special conditions under which media

  aggression can provoke aggressive responses in people, the pattern of findings from

  other studies does not easily lend support to this.  For example Walters and Thomas

  (1963) used the film Rebel Without a Cause and a control film about picture-making

  by teenagers to investigate the effects of film aggression.  Before the experiment the

  two groups were fairly equal in their aggression but following the films the group who

  had seen Rebel Without a Cause became significantly higher on aggression than the

  control group.  The problem with this study is that Walters did not frustrate the

  students to produce anger-arousal and yet demonstrated an enhancement of

  aggression through the aggressive film. 


  Zillman and Johnson (1973) suggested that some of the discrepancies between

  studies may be explained by the arousal caused by aggressive films.  The researchers

  suggested that what should be added is a control condition of no-film to compare the

  effects of seeing a violent film clip.  They used The Wild Bunch compared with a

  neutral film (Marco Polo's Travels) compared with a no-film condition.  Zillman and

  Johnson took physiological measures of the participants in the films showing that the

  neutral film depressed arousal relative to the no-film condition.  This explained the

  results of the aggression measure which in this case was willingness of the

  participants to deliver electric shocks and here there was no difference between the

  violent film and the no-film condition, both of which produced more aggression than

  the neutral film. 


  Further support for the idea that aggressive films produce arousal rather than

  aggression per se is provided by Tannenbaum (1971; Tannenbaum and Zillman

  1975).  In these studies humorous films produced more aggression than control films

  but aggressive films were equally as effective in increasing humorous responses to

  humorous material.  Similarly non-aggressive erotic films, which were arousing,

  increased rewarding behaviour when the learning task required this rather than

  punishment.  Indeed Mueller, Donnerstein and Hallam (1983) reported a 50% increase

  in pro-social behaviour following exposure to a violent programme.  Zillman went on

  to develop a sophisticated theory of media effects based on the idea of excitation

  transfer (eg  Zillman, 1978, 1979, 1982,1991; Bryant and Zillman, 1994). 


  While these developments are of some interest they pose a large question mark over

  the whole idea that the violence content per se of the media creates social problems,

  in that any arousing experience, an exciting football game for example, could under

  the right conditions increase aggressive responding in the observer.  Moreover during

  the 1960s a number of researchers began to produce evidence that what subjects will

  do in a laboratory setting is very different from what they would do in everyday life (see

  for example Adair, 1973:  Friedman, 1967;  Orne, 1969;  Orne and Holland, 1968;

  Westland, 1978). 


  The concern must be that university students participating in film experiments as part

  of their course requirements may have some good idea of what the experimenter is

  hoping to demonstrate and may simply role play in order to produce the results they

  think the experimenter wants.  As Freedman (1986) has observed merely showing an

  aggressive film may imply that the experimenter expects or even desires aggressive

  responses.  Certainly Borden (1975) found aggressive responses were lower when

  female observers were present and increased if observers appeared to have

  aggressive values.  However, whether or not the participants can guess what the

  experimenter is researching does not seem to have produced a consistent pattern of

  findings. Sometimes sophisticated participants appear to comply with what

  experimenters are expecting and in other studies seem to lean over backwards not to

  do so (for example Page and Scheidt, 1971;  Shuck and Piso, 1974; Berkowitz, 1971).



  Experimental Field Studies


  While experiments taking place in more naturalistic settings are not necessarily free

  from such potential artefacts they at least move the study of media effects away from

  university student populations.  The first attempt to apply experimental methods to a

  natural setting was carried out by Feshbach and Singer (1971) in three private schools

  and four boys' homes.  The boys were subjected to either a diet of violent television

  or a diet of neutral television.  Some immediate problems arose:  the neutral group

  complained bitterly about their diet and by popular request the television series

  Batman (a violent programme) had to be allowed.  The experiment continued for six

  weeks, during which house-parents and teachers recorded the boys' behaviour.  In the

  private schools little difference was observed between the violent television group and

  the neutral television group.  However, in the boys' homes aggression towards peers

  in the violent television group was almost half that of the neutral television group.

  Feshbach and Singer concluded this was further evidence for catharsis, but this was

  an unpopular conclusion and stimulated many critical reviews. 


  One of the most serious methodological  problems in this study was that the observers

  knew which programmes the children had watched.  Wells illuminated this problem

  (Wells, 1973) in a similar field experiment where he manipulated the television diet of

  young people in 10 residential schools.  Overall the main finding was that verbal

  aggression was higher in the neutral television group than in the violent television

  group, thus supporting Feshbach.  However, Wells noted some causes for concern:

  there was no significant effect overall on physical aggression but boys who were

  above average on aggression in the violent television group became more aggressive

  during the experiment.  However, this finding was associated with those observers

  who inadvertently had become aware of the group to which a child belonged.  On

  close analysis it seems that the boys were more likely to be rated as aggressive if they

  were known to be watching violent films. 


  Probably the most unusual and adventurous series of field experiments were carried

  out by Milgram and Shotland (1973) who carried out a sustained series of experiments

  attempting to provoke anti-social behaviour through the use of  films.  In the original

  experiment over 600 adult subjects were recruited to attend a film screening in a

  National Television Preview theatre in New York, with a promise that they would

  receive a free radio for attending.  They saw one of three films all starring the hero

  Tom Desmond who loses his job and finding himself desperate for money tries to get

  it back only to find that his services are no longer required.  When he sees his boss

  participating in a telethon to raise funds for a community clinic Tom first dials the

  telethon and pours abuse down the telephone at his boss and then smashes the

  telethon collection box and pockets the contents and runs into the street.  The film was

  shown with various endings: in one version Tom is caught by the police and sent to

  jail and it appears that his marriage is breaking up (negative ending); in another he

  escapes to Mexico where his wife plans to join him (successful outcome); and the third

  version (neutral version) Tom thinks about breaking into the collection box but

  restrained by thoughts of  his family contributes a coin instead.  Finally a neutral non-aggressive  film was shown which dealt with the love affair of a diplomat. 


  After seeing the film participants were asked to go to a different building in order to

  collect their radio.  They arrived one by one to an empty office to find a notice on the

  door saying that no more radios were being distributed.   In the room was a charity

  collection box similar to the one seen in the film, with a dollar note protruding from it.

  Of course the reactions of the participants were being observed from behind a one-way screen.  Most of the subjects became irritated, paced around the room, looked at

  the collection box and left within a few minutes.  However, nearly one in five found the

  temptation too hard to resist:  15 broke the box and took the money from it;  10

  removed the dollar note; 20 tried unsuccessfully to break into the box and 31 people

  stole other things.  However the important point is there were no differences between

  the different film groups.  The proportion of dishonesty was the same whatever film

  they had seen. 


  In an even more ambitious study they were able to access a TV programme  Medical

  Centre which they edited to produce two anti-social versions and one pro-social

  version of the same story.  This last series of field experiments, which had audiences

  of around one million people, again they found no evidence of imitative anti-social

  behaviour due to the films. 


  The only important studies using experimental field techniques to conclude

  unreservedly that films mediate aggression have been associated with Leonard

  Berkowitz (Leyens, Camino, Parke and Berkowitz, 1975;  Parke, Berkowitz, Leyens,

  West and Sebastian, 1977).  One study took place in America and the other in

  Belgium.  Both field experiments used institutions for secondary school boys who

  lacked either adequate homecare or had got into trouble of some sort or another.  The

  boys were observed for one week before the experiment began, and then during the

  week of the experiment in which they watched commercially produced films rather than

  television shows.  Five of the films were aggressive, such as Bonnie and Clyde and

  The Dirty Dozen and five were neutral or comedy or family films. 


  In common with many previous studies by Berkowitz the results seem very impressive

  at first sight.  For example the effects of one or more violent films was to increase

  aggression in the young boys by more than forty fold.  This is really quite curious since

  during the film experiments the young people were not allowed to watch their normal

  television.  Why should films be so dramatically more effective than television?

  However whatever second thoughts might be on the results there is no escaping the

  serious weakness in the research.  It seems that each boy was observed at the most

  for a total of three minutes a day, so the level of measurement was fairly cursory.

  More than this, on close examination of the coding procedures would not seem

  capable of discriminating between a serious fight and active games of basketball,

  since " . . . playful as well as malicious attacks were scored identically . . . ".  Moreover

  the coders responsible for the data collection were all undergraduate students who

  may have been aware of the experimental conditions which they were monitoring,

  although no evidence is available on this point.  Cook, Kendzierski and Thomas

  (1983) conclude that as a body of evidence the field experiments are un-interpretable.





  Given the sophistication of so much research in this field and the obvious identification

  of confounding variables in any link between media violence and real-life aggression

  it is surprising that so many reports have been produced claiming that delinquents

  enjoy aggressive television.  These date back to Haines (1955) who reports that 100

  teenagers in Chicago jails said they thought that films and pornography had

  contributed in some way to their criminal histories.  Essentially similar arguments were

  put forward by Bailey (1993) who suggested that in a large sample of incarcerated

  offenders violent film themes were the most popular fare. 


  As has been noted earlier other factors, such as social class, may provide the causal

  explanation of  both delinquency and preferences for aggressive television fare.  This

  was demonstrated very effectively by Halloran, Brown and Chaney (1970) in a large

  study of 334 known delinquents with two control samples who had not been in trouble

  with the law.  One control group was of 144 working class youngsters matched in

  terms of age, sex, socio-economic status, intelligence and school attainment, and the

  second of 185 boys and girls obtained from slightly higher socio-economic background

  and higher school attainment. 


  All of the participants were interviewed to discover by direct and indirect methods

  whether there were any differences in the importance they attached to television or its

  prominence in their lives.  In table after table there are significant differences between

  the delinquent group and the middle class control samples.  In table after table there

  are significant differences between the two control samples where working class

  youngsters prefer aggressive television programmes compared with the middle class

  control group.  However, in table after table there are no differences between the

  delinquent group and the working class controls.  The researchers concluded that

  television does not have criminogenic effects and that differences in the use of

  television may be accounted by social class more than anything else.


  Nobel in 1971 reported an interesting experiment using repretory interviews

  (structured, probing interviews) to examine the identification preferences of

  delinquents compared with a control group.  The surprising conclusion that he drew

  was that delinquents were less likely to identify with television characters than were

  the control group.  However Nobel puts forward a somewhat dubious argument that

  because delinquents identified less with family members than the control group

  therefore the salience of television characters may be greater. 


  In 1994 the Policy Studies Institute in London published their study of Young

  Offenders and the Media (Hagell and Newburn, 1994).  From a large sample of

  delinquents they selected 200 young people who had been charged or cautioned by

  the police a minimum of three times in one calendar year.  However they were not

  entirely successful in obtaining interviews from all members of this group.  When the

  research closed they had obtained interviews with only 78 young offenders.  The low

  (39%) response rate in the delinquent sample is not very surprising  (West and

  Farrington, 1977).  It probably reflects the chaotic, itinerant lifestyle of young offenders

  more than their unco-operative anti-social attitudes.  Despite the smallness of the

  sample the researchers were clearly successful in identifying a group of evident

  delinquents.  They report that the average number of arrests per offender in 1992 was

  4.5 while the total known or alleged offences in 1992 was 10.8 per offender.  The

  responses of these young offenders to questions about their media use, their preferred

  television programmes, their favourite characters on television and so on were

  compared with a sample of  538 school children of similar age. 


  The results were surprising.  Overall the offenders had less access to television, video

  and other equipment than the comparison school children.  Delinquents as a group

  reported having fewer television sets, fewer video recorders and less access to non-terrestrial

  broadcasting in the places in where they were living. A lower proportion of

  offenders than the comparison school children were able to name any favourite

  television programmes.  The offenders had difficulty in identifying anyone on television

  they would like to be like. However with all the groups, offenders and school children,

  Arnold Schwarzenegger was a popular choice.  For both groups The Terminator. was

  in the top five favourite films.  However cinema attendance was lower in the offender

  group than in school children. Thus half of the offenders said they rarely or never went

  to the cinema compared to a quarter of the school children. 

  The researchers looked more closely at offenders who had been convicted of violent

  offences and found they did not differ in their viewing habits from the group as a

  whole.  Of the offenders who had a favourite television programme the most popularly

  mentioned was the police television series The Bill.  In addition they enjoyed the same

  kind of programmes as the school children including EastEnders, Neighbours, Home

  and Away and Prisoner Cell Block H. 



  Whilst this research does nothing to support concerns about the relationship between

  the media and delinquent offending the results were surprising to many.  The

  researchers had failed to match the group in terms of social class.  Working class

  social groups typically watch almost twice the amount of television and violence as

  their middle class counterparts (Day and Cowie, 1990).  Remembering Halloran,

  Brown and Chaney's study this would seem to have introduced an in-built bias in

  favour of finding that working class young offenders preferred aggressive television

  fare.  However the most obvious conclusion to be drawn from this research echoes

  that of Messner (1986): delinquents are not at home watching television, whether

  violent or not.  They are out on the streets where their peer group socialisation into

  delinquency is the most powerful factor encouraging their offending behaviour (Graef,

  1993).  Of course it remains possible that among such offenders violent videos are

  more of a peer group phenomenon (eg Vogelsang, 1991) but their salience seems

  relatively muted. 


  The most recent research on young offenders is reported by Bayliss (1995) at the

  Cambridge University Institute of Criminology.  This is a thoughtful study focusing on

  the fantasy experiences of young offenders compared with a matched comparison

  group of socially rewarded young men.  Among Bayliss's main findings were firstly that

  in their early teens, or beforehand, many young offenders had embarked on an

  interwoven mental diet of their own fantasies and consumable drugs which together

  " . . . served primarily as analgesics for the chronic emotional pain which had almost

  always emerged in a discordant family atmosphere, and for which they had not

  developed alternative coping strategies . . . ". Secondly, for most young offenders their

  favourite movie character and storyline was seen as reflecting rather than fuelling their

  motivations, " . . . it is Arnie Schwarzenegger in Terminator 2 a robot - living tissue

  over a metal frame - an invisible man who cannot feel pain who shows no emotion in

  voice or face and crucial to their subconscious need a man who travels back in time

  to protect, befriend and die for a fourteen year old boy who has no father . . . "

  (Bayliss, 1995, personal communication).  Unfortunately this kind of approach which

  attempts to understand human imagination and how it relates to the narrative forms

  of film and television is quite remarkably rare in the vast literature on media violence.









  Content and Control


  There is little doubt that people are concerned about crime and violence.  As a cause

  for concern crime has reliably been rated higher than any other issue except

  unemployment in many countries (Watkins and Worcester, 1986).  However crime and

  violence remain quite rare personal experiences for the vast majority of the public.

  Thus according to victimisation surveys most people can look forward to a decade or

  more of crime-free existence and even then the offence is likely to be quite trivial

  (Mayhew et al, 1993). 


  However secondary experiences of serious crime and violence through the mass

  media are a daily occurrence.  Indeed it is quite easy to scare monger about the extent

  of media crime and violence.  Frederic Wertham, an American psychiatrist, claimed

  to have studied adolescents ". . . who in comic books, movies and TV have seen more

  than 10,000 homicides . . " (Wertham, 1954).  More than this, depending on the

  newspapers read, the radio stations listened to and the television programmes

  watched, the average person's secondary experience of crime and violence might run

  to 7,000 acts per annum in the UK and even more than this in other countries

  (Cumberbatch, 1989).


  Attempts to quantify the amount of violence on television have been sporadic.  Thus

  direct comparison between different countries are problematic especially  since the

  time periods sampled vary from one study to another, while eccentric decisions seem

  to have been taken on which programme genres should be included or excluded.

  Moreover definitions of violence vary somewhat depending on whether such things as

  verbal threats or natural disasters are included.  However rates of around 5-6 violent

  acts per hour seem to be typical of most countries studied where typically some two

  thirds of programmes contain some violence (Cumberbatch, Jones and Lee, 1988).

  Such rates were recorded for Australia, New Zealand, France, The Netherlands and

  the USA.  Somewhat higher rates were observed in Germany and Japan while Finland,

  Great Britain, Sweden and Israel were much lower.


  The focus of most content analyses has been on terrestrial channels where public

  service broadcasting has become eroded over the years (McKinsey, 1993) and been

  supplemented by a growing number of cable and satellite channels.  Comparisons

  between terrestrial and cable/satellite are not easily made since most terrestrial

  channels offer a range of programme material within each channel while cable/satellite

  achieve a variety across channels.  However movie channels might well be expected

  to contain more violence than typical of terrestrial television as confirmed by recent

  monitoring by the Broadcasting Standards Council in the UK.  In this study three

  satellite movie channels averaged rates of violence of 7-8 acts per hour which is

  almost double that of terrestrial television (4 violent acts per hour) over the same two

  week period (Cumberbatch, Maguire and Woods, 1994).  An additional finding which

  must be relevant to most countries is that programmes from the USA contained around

  twice as much violence (and indeed bad language) as UK productions.  This was

  examined in some detail in an earlier study for the BBC which analysed 1,412 hours

  of television output.  Here American productions averaged 3.6 violent acts per hour

  compared with only 1.1 acts per hour for British programmes.  The authors attribute

  the difference to the greater predominance of violent genres such as dramatic fiction

  and cartoons in the American programmes.  However even when these genres are

  compared American productions still contained more violent acts at 3.8 per hour

  compared with 2.5 per hour for British productions (Cumberbatch et al, 1987).


  These data suggest that the growth of the cable/satellite coupled with the decline in

  public service broadcasting might well lead to a steady growth in the proportion of

  productions from the USA and consequent increases in the amount of violence.

  Trends over time have not been reliably monitored but Viemero (1990) noted

  considerable increases in violence on Finnish television over the previous decade,

  while British and American research suggest little has changed.


  What is certainly the case is that the growth of television channels allows the

  possibility of children watching a great deal more violence.  However the availability

  of more choice in most European countries has not had a corresponding increase in

  the total amount of time people spend watching television.  On average Europeans

  spend 2-3 hours per day watching television and total exposure has remained fairly

  consistent for the last two decades (Wartella, 1995).


  The most striking phenomenon in audience research data is the considerable social

  class variation noted in many countries where middle class groups watch far less

  television than lower socio-economic groups.  Thus in the UK professional/managerial

  groups average just over 17 hours per week compared with nearly 31 hours for semi-skilled and unskilled manual occupational groups (Day and Cowie, 1990).  Similar

  findings have been observed in Germany (Lukesch, 1987) and Sweden (Rosengren

  and Windahl, 1989). 


  Age differences in viewing are more modest.  For example 4-7 year olds averaged 149

  minutes per day, 8-11 year olds watched 154 minutes, 12-15 year olds rose to 166

  minutes, and 16-24 year olds reduced to 149 minutes (Broadcasters' Audience

  Research Board, 1990).  Similar trends have been noted in Sweden, Finland, Belgium

  and Norway (Wartella, 1995).


  Despite the popularity of violent action adventure films, Wartella (1995) suggests that

  soap operas, comedy and game shows feature in the top ten in all European countries

  for which such charts are available.


  However  the television diet enjoyed by young people changes with age.  Very young

  children (under the age of six) mainly watch programmes made for child audiences

  whereas older children (10-12 years old) tend to watch very little children's

  programming (Gunter, 1987).


  Part of the reason why soap operas, game and comedy shows are popular with young

  people is that television viewing remains a family entertainment.  Here television forms

  a pivotal role with family routines structured around television habits (Alexander,

  1994).  Evidence is somewhat ambiguous on the extent to which child audiences

  modify family viewing patterns.  However by and large programme content does not

  seem to be the major factor in assembling an audience for television - the decision

  whether or not to view television takes precedence over what to view (Barwise and

  Ehrenberg, 1988).  Thus television is often described as a "low involvement medium"

  where viewers generally are indifferent to its content (Comstock, 1991).


  Few children report restrictions on what they watch even though nearly nine out of ten

  adults (88%) in one survey believed that ". . . parents should take greater care over

  the TV programmes their children watch at home . . ." (Gunter, 1988).  Despite the

  apparent absence of rules on children's viewing, survey data suggests that families

  are more likely to watch action adventure programmes with older children (11-15 year

  olds) than younger ones.  Moreover the proportion of parents watching any particular

  action adventure with their children is directly related to how "harmless" it is rated in

  surveys.  Thus seven out of ten parents watched The A Team with their children while

  fewer than one in ten watched Hill Street Blues with their children present (Gunter,



  Conventionally content analyses would record The A Team as more violent than Hill

  Street Blues.  This reveals the weakness in so many attempts to monitor the amount

  of violence on television.  Many studies have recorded appalling rates of violence in

  children's programming due to such cartoons as Tom and Jerry.  However surveys of

  public perceptions of television violence reveal that such shows are not perceived to

  be violent since "violence" contains a realism element (Howitt and Cumberbatch,

  1974;  Gunter, 1985).  For this reason it would seem crucial for content analyses to

  monitor television violence in terms of the programme genres.  Cumberbatch et al

  (1987) categorised The A Team as "real life cartoon" and focused on violence in

  contemporary realistic drama for their discussion of violence which would be reflected

  in public concerns.


  Realism and graphic detail are certainly among the most prevalent forms of violence

  to produce fright responses and are common.  Blumer (1933) reported that 93% of the

  children in his sample had been frightened by a motion picture.  Wilson, Hoffner and

  Cantor (1987) report that 75% of their sample had been scared by something they had

  seen on television or in a movie.  However they also found that over two thirds of

  children said they enjoyed scary programmes!


  This is of course one of the central issues in debates about media violence.  Sadly

  many people enjoy violence in the media and children do not differ from their parents

  where 75% prefer entertainment programmes to educational ones.  This represents

  a challenge for those who wish to improve television for children (Stipp, 1993).


  Studies of viewing behaviour and experiences have been limited compared with the

  research effort put into experiments designed to explore the effects of television.

  However they introduce a challenge to many assumptions such as that children are

  glued to the box passively absorbing anti-social messages.  From these studies it

  seems clear that television viewing by children remains largely a social event but can

  serve an almost unlimited range of diverse uses and functions.  They can use it to be

  with the family or to get away from it, to stimulate conversation or avoid it and so on

  (Alexander, 1994).


  Perhaps most importantly half of the time that the television set is on the audience is

  doing something else.  Talking is the most common other activity (Gunter and

  Svennevig, 1987).  Moreover even young viewers interpret what they see according

  to their experience and knowledge of the world such that television is a weak tool to

  undermine the beliefs and values of young people (Buckingham, 1993; Hodge and

  Tripp, 1986).  Indeed many of the assumptions underlying concerns about media

  violence are fuelled by an oversimplified view of both television content and the

  viewing audience.


  Perhaps the most pernicious of these is the failure to discriminate between the rich

  variety of television fare by categorising all programmes with violence in them as

  "violent" and therefore bad.  Much dramatic fiction revolves around deeply moral

  narratives where good triumphs over evil.  The heroes of "violent cartoons" like

  Batman are overtly and even excessively moral beings.  Indeed the pattern of

  television violence is typically that violence for personal gain is punished and only

  rewarded when used by "goodies" (forces of law and order) to defeat "baddies"

  (Cumberbatch et al, 1987) a point appreciated by even four year old viewers

  (Buckingham, 1993).


  One final point to be mentioned of the content of television and audience diets was

  noted by Wiegman et al (1992).  Heavy viewers of violent programmes were also

  heavy viewers of pro-social programmes.  Whether the failure to find any influence of

  violent television on children's aggressivity in this study was due in anyway to their

  exposure to pro-social programmes is impossible to say but should not be neglected

  in future research.






  This review has focused on some of the methodological problems with research in this

  field.  Investigations of the relationship between television violence and aggression

  can too easily contain an in-built bias in favour of finding some link.   For example in

  correlational studies which tend to predominate it is inevitably the case that children

  who are rejected by parents and abused by them are likely to become aggressive and

  may seek television as a fantasy world to escape from the harsh realities of their day-to-day existence.  The extent to which television plays a role in their development has

  not been teased out adequately in existing research. 


  Similarly with the experimental studies there seems  little doubt that the arousal

  produced by watching action, fast-paced aggressive programmes may be the primary

  source of immediate aggressive responses, as measured in so many studies, but

  whether this is an artefact of the testing procedure or has any lasting effect seems in



  An overriding consideration in terms of methodology is the tradition of scientific

  research to report significant findings and not handle non-significant ones.  This leads

  to a particular concern in this field that so many studies may be torturing the data until

  it confesses something.  The fact that these confessions are different in different

  studies may remind us of the unreliability of confessions extracted under torture.

  Scientists may be reluctant to consider non-significant findings and indeed they are

  rarely  part of any scientific debate and are unlikely to be published at all (Greenwald,

  1975;  Melton, 1962;  Smith, 1980;  Westland, 1978).  However especially in the

  correlational studies which take so many measures the vast data show up a

  prominence of non-significant findings which should at least alert us to the danger that

  those that are reported may in fact be unreliable. 


  It is worth mentioning in this context that the massive study by Belson found no

  significant effects of television violence on boys' respect for authority; on their

  consideration for other people; on their sleep disturbance or on stealing (Belson,

  1975).  Violent programme preferences were not linked to their preoccupation with

  acts of violence shown on television, or youngsters feeling more willing to commit

  those acts or to see violence as a basic part of human nature or to accept violence as

  a way to solve their problems.  Simple summaries of such enormously detailed

  research findings focusing on one measure which offers a statistically significant link

  is thus quite misleading.  It is also worth remembering that Belson found equally strong

  relationships between newspaper readership and aggressive behaviour as he did

  between television violence and aggression. 


  In many ways the persistent problems in this field are more conceptual than

  methodological.  It is difficult to avoid the feeling that most studies are designed to

  prove that television violence causes aggression rather than to understand the role

  of television in the social development of children.  Children who watch a lot of

  television will watch a lot of violence and a lot of other television fare which is overtly

  pro-social.  It is true that over the years more attention has been paid to possible pro-social effects of television (eg Liebert and Sprafkin, 1988; Liss, Reinhardt and

  Fredriksen, 1983;  Rushton, 1979; Sprafkin, Gadow and Kant, 1988; Wiegman,

  Kuttschreuter and Baarda, 1992).  However until an evaluation is made of the role of

  television in all aspects of children's lives and their socialisation it may be risky to

  draw conclusions about one particular aspect of television's output.  This is particularly

  important in the context of the predominant theories and methods which are so

  influenced and informed by psychological traditions.  Until recently these had been

  remarkably crude with the inevitable consequence in terms of the research designs

  adopted.  It would seem essential that research in this field becomes informed by

  current theorising within the fields of criminology and of media studies. 


  At the theoretical level the notion that violence has a simple uni-dimensional effect on

  the viewer has been challenged for a long time by the growing tradition of " uses and

  gratifications" research which seeks to identify the dynamic relationship that particular

  individuals with particular needs and world views may enjoy with the mass media (eg

  Bryant and Zillman, 1991; Palmgreen and Rayburn, 1985; Rosengren, Wenner and

  Palmgreen, 1985; Taylor and Mullan, 1986; Zillmann, 1985).  Essentially this research

  perspective demonstrates that viewers are not passive in their use of the media but

  actively bring to the viewing situation their own particular attitudes and values which

  mean that what they take away from the viewing experience will depend on those



  A further conceptual weakness in the research literature to date is a failure to consider

  the history and politics of concerns about media violence (eg Rowland, 1983).  As

  suggested earlier there is a strong tradition of concerns about popular culture.  All too

  easily academics may be seen as those in opposition to popular culture.  The very

  long history of concerns about rising crime and the extent to which each new medium

  introduced may aggravate this is a matter that is worthy of further attention.  Does the

  mass media coverage of such concerns influence research perspectives on the

  problems?  Evidence for this is circumstantial but has been strongly argued by

  Cumberbatch (1994) in a critical review of a report produced by Newson (1994)

  commissioned by a Member of Parliament  to support his campaign to introduce

  additional controls over the sale and distribution of videos (Home Office, 1994).  The

  report by Newson relies heavily of newspaper accounts of rising crime, especially

  among juveniles, to assert its case that we have a new problem which must be caused

  by violent videos.  Press speculation that violent videos caused two ten year old boys

  to murder two year old James Bulger was rife at the time.  However neither in this case

  (Smith, 1993) nor in previous controversial cases (eg Josephs, 1993) has any solid

  evidence been forthcoming.


  There would seem little doubt that criminal statistics do suggest a massive rise in

  violent crime.  However it is also clear from victimisation surveys carried out in the

  United States and the United Kingdom that the growth in experienced violent crime is

  small compared with the growth in criminal statistics.  In other words as Mayhew et al

  (1993) argue more people increasingly report violent crime but the actual probability

  of criminal victimisation lags well behind this.  For example Mayhew points out that the

  100% increase in recorded violent crime in the UK over the last decade is matched by

  a mere 15% increase in actual victimisation.  There is an obvious point that the role

  of the mass media in this should be considered.  More people are concerned about

  violent crime and this in itself may fuel fears about the mass media.  Clearly we need

  to know much more about the way in which the mass media may shape our

  perceptions in society, not just at the level of encouraging anti-social behaviour but

  encouraging strong feelings against it.  In this context it is worth mentioning the

  contribution of George Gerbner to this debate.  He has argued that the role of

  television is to encourage a fear of crime in society.  Though evidence on this is

  controversial (Gunter and Wober, 1987) it is something that must be part of any

  consideration of the power of television (Gonczol, 1993). 


  This point leads to broader considerations about the role of the mass media in society.

  Almost all of the research literature focuses on the simple question "Does mass media

  violence cause individuals to be aggressive?".  In the criminology literature this idea

  of individuals responding uniquely to a particular factor is not taken seriously.  Indeed

  the stability of aggression is quite remarkable whereby anti-social personalities seem

  to be formed as early as five years old (Olweus, 1980; Robins, 1991; Smith, 1995).

  Moreover delinquents are not getting younger - in the UK the crime rate peaked at 13

  in 1938, 14 in 1961 and 15 in 1983 (Farrington, 1986).  Similar findings are reported

  in Sweden by Wikstrom (1990).  These data rather undermine the thesis that television

  has achieved some impact.  The issue of crime in society is one that is much broader

  than just offenders - importantly it also involves victims who may or may not be less

  vulnerable because of what they see on television. Fear of crime for example which

  might be aggravated by television may encourage more prudence in dealing with

  potential criminal assaults (Mendelsohn, 1983).  The same thing may apply to the

  social processes of producing crime.  Are witnesses more or less likely to report anti-social behaviour or even intervene in witnessing such scenes because of violence on

  television.?  And what are the effects on the police, the judiciary or the legislature or

  indeed public perceptions of these (Cumberbatch and Morgan, 1985)?  Mayhew

  (1993) has argued that the increase in criminal statistics over the victimisation rate in

  society is due mainly to the police taking more active interest in the reporting of violent

  crime.  Any approach to understanding aggression or crime should consider the

  processes which "cause" crime and alter the criminal statistics. 


  The absence of policy orientated research is a particular weakness.  For example

  broadcasters need to know what kind of violence should they curtail.  Is it fictional

  violence set in a contemporary setting?  Or is it violence in historical drama? Is it

  violence in fictional programming set in the future?  Is it important whether or not to

  show graphic detail to repulse viewers and make them understand the consequences

  if they commit acts of aggression or is it better to remove such details?  Is it possible

  to advise what kind of programmes might replace those that contain violence, for

  example is Sesame St much less harmful than Power Rangers?  Are the programmes

  which feature violence in cartoons more dangerous than more realistic productions?

    Unfortunately answers to these questions are simply not available from the research

  evidence (see Dorr, 1988).


  In general terms the research literature on mass media violence is disappointing.  in

  the lack of an overall game plan informed by relevant disciplines.  More than anything

  it lacks an applied focus  which could help inform governments, broadcasters and

  producers on what they should do. 


  However in the Pandora's box of research there is one practical solution that stands

  out.  A number of studies have carried out interventions attempting to persuade young

  people that television violence is not a good thing and it should be viewed critically,

  for example Huesmann, Eron, Clean, Price and Fisher (1983) asked school children

  to write essays on why television violence is bad and unreal.  This not only reduced

  aggression as rated by peers but reduced that correlation between violence viewing

  and aggression which has been so much the focus of so much research.  Similar

  experiments are reported by Singer, Zuckermann and Singer (1980);  Anderson,

  (1983);  Sprafkin, Swift and Hess (1983); Sprafkin, Gadow and Kant (1988);  Eron

  (1986); and Vooijs and Van der Voort (1993). 


  Such interventions may not tell us anything about the usual effects of television and

  may not even have anything to with the effects of television per se but it would seem

  the most promising line to follow.  After all given the growth of satellite and cable and

  other systems to reproduce film material the violent media horse has already bolted.

  Thus the most practical solution to the issue of media violence must be to introduce

  media education into the curricula of all countries.  This could not do any harm and

  would at least inform debate on this vexed issue.  Finally, the question which has not

  been raised so far in the research literature is why so many people are not affected

  by television violence.  If we can begin to understand how people can protect

  themselves against such potential influences then we may begin to develop an

  understanding of how people can cope with all vicissitudes in life and develop social

  programmes for this.





  ABELSON, R.P. (1976) 'Script processing in attitude formation and decision making' in: J.

  Carroll and  J. Payne (eds) Cognition and Social Behaviour, Hillsdale, N.J: Lawrence



  ADAIR, J. G. ( 1973) The Human Subject: the social psychology of the psychological

  experiment Boston, Mass,Little: Brown


  ALEXANDER, A. (1994) 'The effect of media on family life' in D. Zillman, J. Bryant and

  J. Huston (eds) Media, Children and the Family, Hillsdale, NJ: Lawrence Erlbaum.


  ANDERSON, J. A. (1983) 'Television literacy and the critical viewer' in: J. Bryant & D. R.

  Anderson (eds). Children's Understanding of  Television  New York: Academic Press pp.



  ANDISON, F.R. (1977)'TV violence and viewer aggression: a cumulation of study results

  1956-1976' Public Opinion Quarterly 41(3), pp. 314-331.


  BAILEY, S. (1993) 'Fast forward to violence' Criminal Justice Matters (3), pp.6-7.


  BANDURA, A. ( 1973) Aggression: A Social Learning Analysis Englewood Cliffs, New

  Jersey, Prentice-Hall.


  BANDURA, A. (1983) 'Psychological mechanisms of aggression' in: R. G. Green and C. I.

  Donnerstein (eds) Aggression, Theoretical and Empirical Reviews: Vol I Theoretical and

  Methodological Issues New York: AcademicPress, pp. I-40.


  BANDURA, A. (1986) Social Foundations of Thought and Action: A Social Cognitive

  Theory Englewood Cliffs: Prentice-Hall.


  BANDURA, A. (1994) 'Social cognitive theory of mass communication' in: J. Bryant and

  D. Zillman Media Effects: Advances in Theory and Research, Hillsdale, NJ: Lawrence

  Earlbaum, pp.61-90


  BANDURA, A. and HUSTON ALETHA, C. (1961) 'Identification as a process of incidental

  learning' Journal of Abnorrnal and Social Psychology 63, pp. 311-318.


  BANDURA, A., ROSS, D. and ROSS, S. A. (1961) 'Transmission of aggression through

  Imitation of aggressive models' Journal of Abnormal and Social Psychology 63, pp. 575-



  BANDURA, A., ROSS, D. and ROSS, S. A. (1963a) 'Imitation of film-mediated aggresive

  models'  Journal of Abnorrnal and Social Psychology 66, pp. 3-11.


  BANDURA, A., ROSS, D. and ROSS, S. A. (1963b) 'Vicarious reinforcement and imitative

  learning' .Journal of Abnormal and Social Psychology 67, pp. 601-607.


  BARKER, M. (1984a) A Haunt of Fears. London: Pluto.


  BARKER, M. (1984b) (ed.) The Video Nasties London: Pluto.


  BARKER, M.(1989) Comics, Influences, Ideology  Manchester: Manchester University



  BARLOW, G. and  HILL, A. (1985) Video Violence and Children  London: Hodder &



  BARON, L. and STRAUSS, M. (1989) Four Theories of Rape: a State Level Analysis, New

  Haven, Conn: Yale University Press.


  BARON JONES, N.  and  REISS, P S. (1985) 'Same time, next year: Aggregate analyses of

  the mass media and video behavior'  American Sociological Review 50, pp. 347-363.


  BARWISE, T.P. and EHRENBERG, A.S.C. (1988) Television and its Audience Newbury

  Park, Calif: Sage


  BAYLISS, N.V.K. (1995) Reality, Fantasy, Fiction and Offending  Unpublished M.Phil.

  Dissertation, Institute of Criminology, University of Cambridge.


  BELSON, W. A. (1975) Juvenile Theft: the causal factors  London: Harper & Row.


  BELSON, W. A. (1978) Television Violence and the Adolescent Boy



  BELSON, W. A.(1985) 'The effects of television violence'  Bulletin of the British

  Psychological Society 38,  pp.120-122.


  BERKOWITZ, L. (1962)  Aggression: A Social Psychological Analysis New York:



  BERKOWITZ, L. (1968) 'Impulse, aggression and the gun'  Psychology Today 2(4), pp.l8-



  BERKOWITZ, L.(1971) 'The "weapons effect", Demand charactcristics and the myth of the

  compliant subject' Journal of Personality and Social Psychology 20(3), pp.332-338.


  BERKOWITZ, L. (1984) 'Some effects of thoughts on anti- and prosocial influences of

  media events: A cognitive-neoassociation analysis'  Psychological Bulletin  95(3), pp. 410-



  BERKOWITZ, L. (1986) 'Situational influences on reactions to observed violence'  Journal

  of Social Issues  42(3), pp. 93-103.


  BERKOWITZ, L. (1993) Aggression: Its Causes, Consequences and Control  New York,

  N.J: McGraw Hill.


  BERKOWITZ, L. & GEEN, R. G. (1966) 'Female violence and the cue properties of 

  available targets'  Journal of Personality and Social Psychology 3, pp.525-30.


  BERKOWITZ, L. & GEEN, R. G. (1967) 'Stimulus qualities of the target of aggression' 

  Journal of Personality and Social Psychology 5, pp.364-  368.


  BERKOWITZ, L. & LE PAGE, A. (1967) 'Weapons as aggression-eliciting stimuli'  Journal

  of Personality and and Social Psychology 7, pp202-207.


  BERKOWITZ, L. & RAWLINGS, E. (1963) 'Effects of film violence on inhibitions against

  subsequent aggression' Journal of Abnorrnal and Social Psychology 66, pp.405-12.


  BERMAN, A. L. (1988) 'Fictional depiction of suicide in television films amd imitation

  effects' The American Journal of Psychiatry 145(10), pp. 982-986.


  BERRY, G.L. and ASAMEN, J.K. (1993) Children and Television  Newbury Park, Calif:



  BLUMER, H. (1933) Movies and Conduct  New York, NY: Macmillan.


  BORDEN, R J. (1975) 'Witnessed aggression' Journal of Personality and Social Psychology

  31, pp.567-573.


  BRODY, S. (1977) Screen Violence and Film Censorship  London:H.M.S.O.


  BROSIUS, H.B. (1987)' Auswirkungen der Rezeption von Horror-Videos auf die

  Legitimation von aggressiven Handlungen' Rundfunk und Fernsehen, 35, pp. 71-91.


  BROSIUS, H.B. und HARTMANN, T. (1988) 'Erfahrungen mit Horrorvideos bei Schlern

  unterschiedlicher Schultypen.  Eine Umfrage unter 12-bis-15 j„hrigen Schlern

  Communications, 14, 3, pp. 91-112.


  BRYANT, J., CARVETH, R. A.  and  BROWN, D. (1981) 'Television viewing and anxiety' 

  Journal of Communication 31, pp. l06-119.


  BRYANT, J. and ZILLMAN, D. (1991) Responding to the Screen:  Reception and Reaction

  Processes Hillsdale, N.J: Lawrence Erlbaum.


  BRYANT, J. and ZILLMAN, D.  (eds) (1994) Media Effects, Advances in Theory and

  Research Hillsdale, N.J: Lawrence Erlbaum.


  BUCKINGHAM, D.(1993) Children Talking Television  London: The Falmer Press.


  BURKE, P.(1978) Popular Culture in Early Modern Europe  London, Temple Smith.


  BURNE, J.(1993) 'When the medium's message is violent'  New Scientist.


  BUSS, A. H., BOOKER, A., and  BUSS, E. (1972) 'Firing a weapon and aggression' Journal

  of Personality and  Social Psychology 22, pp. 96-302.


  CANTOR, J. (1994) 'Fright reactions to the mass media' in: J. Bryant and D. Zillman Media

  Effects: Advances in Theory and Research  Hillsdale, NJ: Lawrence Earlbaum, pp.213-245.


  CASHMORE, E. (1994) And There Was Television, London: Routledge.


  CENTERWALL, B.S. (1989) 'Exposure to television as a cause of violence' in: G.

  Comstock (ed.) Public Communication and Behaviour Vol. 2, pp. 1-58, New York, N.Y:

  Academic Press.


  CHARTERS, W.W. (1933) Motion Pictures and Youth : a summary  New York, N.Y:



  CLARK, D.G. and BLANKENBERG, W.B. (1971) 'Trends in violent content in selected

  mass media'  in: G.A. Comstock and E.A. Rubinstein (eds) Television and Social Behaviour

  Vol.1 Content and Control Washington, D.C: U.S. Government Printing Office  pp.188-243.


  COATES, B., PUSSER, H.E. and GOODMAN, I. (1976) ' The influence of 'Sesame St' and

  'Mister Roger's Neighborhood' on children's social behavior at preschool'  Child

  Development, 47, pp. 138-144.


  COMSTOCK, G. and  PAIK, H.(1991) Television and the American Child  San Diego,

  Calif: Academic Press


  COOK, T.D., KENDZIERSI, D.A. and THOMAS, S.V. (1983) 'The implicit assumptions of

  Television Research: and analysis of the 1982 NIMH Report on Television and Behaviour'

  Public Opinion Quarterly, 47, pp. 161-201.


  COUGHLIN, E.K.(1985) 'Is violence on TV harmful to our health? Some scholars,a vocal

  minority say no'  Chronicle of Higher Education March, pp. 65-71.


  CRESSEY, P. G. and  THRASHER, F. M. (1933) Boys Movies and City Streets New York:



  CUMBERBATCH, G.(1984)'Sorting out little white lies from nasty pieces of research' The

  Guardian 25 April 1984


  CUMBERBATCH, G.(1987) The Portrayal of Violence on British Television  London:



  CUMBERBATCH, G. (1988) 'TV violence: Too violent?' Broadcast 19 February 1988, p.



  CUMBERBATCH, G. (1989) 'Violence and the mass media: the research evidence' in G.

  Cumberbatch and D. Howitt (eds) A measure of uncertainty: the effects of the mass media 

  London and Paris: John Libbey.


  CUMBERBATCH, G. (1994) 'Legislating mythology: video violence and children' Journal

  of Mental Health, 3, pp. 485-494.


  CUMBERBATCH, G. and BEARDSWORTH, A. (1976) 'Criminals victims and mass

  communications' in: E. G. Viano (ed.) Victims and Society  Washington DC: Visage. pp.



  CUMBERBATCH, G. and HOWITT, D. (1974) 'Identification with aggressive television

  characters and children's moral judgements' in: J.Dewitt and W.W.Hartup (eds)

  Determinants and Origins of Aggressive Behaviour  The Hague: Mouton, pp. 517-523.


  CUMBERBATCH, G. and HOWITT, D. (1975) Mass Media Violence and Society,

  London: Elek; New York NY: Halstead Wiley.


  CUMBERBATCH, G. and HOWITT, D.(1989) A Measure of Uncertainty: the effects of the

  mass media  London and Paris: John Libbey.


  CUMBERBATCH, G., JONES, I. and LEE, M. (1988). 'Measuring violence on television'

  Current Psychological Research and Reviews 7,1, pp. 8-23.


  CUMBERBATCH, G., LEE, M., HARDY, G., and JONES, I. (1987) The Portrayal of

  Violence on British Television, London: BBC.


  CUMBERBATCH, G., MAGUIRE, A. and WOODS, S. (1994) 'Content Analyses'

  Broadcasting Standards Council Monitoring Report III, London: Broadcasting Standards

  Council, pp. 29-230.


  CUMBERBATCH, G. and MORGAN, B. (1985) 'The police officer' in: W. T. Singleton

  (ed.) The Study of Real Skills Vol 4, Lancaster: MTP Press.


  DAY, K. and COWIE, E. (1990) 'Trends in viewing and listening' BBC Broadcasting

  Research Annual Review, XVI, pp. 5-22.



  (1939) Frustration and Aggression New Haven, Conn: Yale University Press.


  DONERSTEIN, E. and BERKOWITZ, L. (1981) ' Victim reactions in aggressive erotic

  films as a factor in violence against women'  Journal of Personality and Social Psychology

  41, pp. 710-724.


  DORR, A. (1988) 'When social scientists co-operate with broadcasters'  in: S. Oskamp (ed.)

  Television as a Social Issue: Applied Social Psychology Annual, Vol 8, Beverley Hills,

  Calif: Sage, pp. 285-303.


  DUHS, C.A. and GUNTON, R.J. (1988) 'TV violence and childhood aggression: A

  curmudgeon's guide' Australian Psychologist 23 (2), pp.183-195.


  ERON, L.D. (1963) 'Relationship of TV viewing habits and aggressive behavior in children'

  Journal of Abnormal  and Social Psychology 67, pp.193-196.


  ERON, L. D. (1982) 'Parent-child Interaction, television violence and aggression in children'

  American Psychologist 37(2), pp. 197-211.


  ERON, L.D. (1986) 'Interventions to mitigate the psychological effects of media violence on

  aggressive behaviour' Journal of Social Issues, 42 (3), pp.155-169.



 'Age trends in the development of aggression, sex-typing and related television habits'

  Developmental Psychology (19), pp.71-77.


  ERON, L. D., HUESMANN, L. R., LEFTKOWlTZ, M. M. and  WALDER, L. O. (1972)

  'Does television violence cause aggression?' American Psychologist 27, pp. 253-263.


  EYSENCK, H. J. and NIAS, D. K. (1978) Sex Violence and the Media London: Maurice

  Temple Smith.


  FARRINGTON, D.P. (1986) 'Age and crime' in: M. Tonry and N. Morris (eds) Crime and

  Justice: An Annual Review of Research, Vol VII  Chicago: University of Chicago Press,



  FELSON, R.B. and MORGAN, M. (1984) 'Book review: Television and aggression: a panel

  study' Contemporary Sociology, 13 (2), pp. 123-125.


  FESHBACH, S. (1955) 'The drive-reducing function of fantasy behavior' Journal of

  Abnorrnal and Social Psychology 50, pp. 3-11.


  FESHBACH, S. ( 1963) 'The stimulating versus cathartic effects of a vicarious aggressive

  activity' Journal of Abnormal and Social Psychology 63, pp.381-5.


  FESHBACH, S. and  SINGER, R. D. (1971) Television and Aggression: An Experimental

  Field Study San Francisco: Jossey-Bass.


  FREEDMAN, J. L. (1992) 'Television violence and aggression: What psychologists should

  tell the public' in: P. Suedfield and P.E. Tetlock (eds) Psychology and Social Policy  New

  York, NY: Hemisphere,  pp.179-189.


  FREEDMAN, J. L. (1984) 'Effects of television violence on aggression' Psychological

  Bulletin 96(2), pp.227-246.


  FREEDMAN, J. L. (1986) 'Television violence and aggression: a rejoinder' Psychological

   Bulletin 100(3), pp.372-378.


  FREEDMAN, J. L. (1988) 'Television violence and aggression:what the evidence shows' in:

  S.Oskamp (ed.) Television as a social  issue.  Applied Social Psychology Annual No 8 ,

  Newbury Park, Califomia: Sage.  pp.144-162.


  FRIEDMAN, N. (1967) The Social Nature of Psychological Research: The Psychological

  Experiment as a Social Interaction New York: Basic Books.


  FRIEDRICH-COFER, L. and  HUSTON, A. C. (1986) 'Television violence and aggression:

  The debate continues' Psychological Bulletin 100 (3), pp.364-371.


  FRODI,  A. (1975) 'The effect of exposure to weapons on aggressive behavior from a cross-

  cultural perspective' International Journal of Psychology 10, pp.283-292.


  FROSCHHAMMER, H. (1992) Medienkonsum und Delinquenz.  Ein Vergleich ber

  Wirkungsprozessse bei Heimkindern und bei Kinden aus Familien, Regensburg:

  Unver”ffentlichte Diplomarbeit.


  FRYDMAN, M. (1995) 'Violence, vandalism and aggressiveness' Journal of Environmental

  Psychology, Toxicology and Oncology (In press).


  GADOW, K. D. and SPRAFKIN, J. (1989) 'Field experiments of television violence'

  Pediatrics 83, pp.399-405.


  GADOW, K. D. and SPRAFKIN, J. (1993) 'Television violence and children with

  emotional and behavioral disorders' Journal of Emotional and Behavioral Disorders, 1, (1),

  pp. 54-63.


  GAUNTLETT, D. (1995) Moving Experiences: Understanding Television's Influences and

  Effects London: John Libbey.


  GEEN, R.G. (1994) 'Television and aggression: Recent developments in research and

  theory' in: D. Zillman J. Bryant and A.C. Huston (eds) Media, Children and the Family: 

  Social Scientific, Psychodynamic and Clinical Perspectives  Hillsdale, N.J: Lawrence

  Erlbaum, pp.151-162.


  GEEN, R.G. and  O'NEALL, E. (1969) 'Activation of cue-ellicited aggression by general

  arousal' Journal of Personality and Social Psychology 11(3), pp. 289-292.


  GERBNER, G. (1972) 'Violence in television drama: trends and symbolic functions' in: G.

  A. Comstock and E. A. Rubenstein (eds) Television and Social Behavior,Vol 1, Media

  Content and Control  Washington DC: US Govemment Printing Office. pp.28-187.


  GERBNER, G. (1990) 'Epilogue: Advancing the path of righteousness (maybe)' in N.

  Signorielli and M. Morgan (eds.) Cultivation Analysis: New Directions in Media Effects

  Research, Newbury Park, Calif: Sage.



  and SIGNORIELLI, N.(1977) 'Television violence profile No 8.' Journal of Communication

  27, pp.171-180.


  GLOGAUER, W. (1990) 'Fallbeispiele medieninduzierter Delinquenz'  in: H Lukesch

  (Hrsg.) "Wenn Gewalt zur Unterhaltung wird ... " Beitr„ge zur Nutzung und Wirkung von

  Gewaltdarstellungen in audiovisuellen Medien, Regensburg: S. Roderer, pp. 149-159.


  GLOGAUER, W. (1991) Kriminalisierung von Kindern und Jugendlichen durch Medien.  

  Wirkungen gewaltt„tiger, sexueller, pornographischer und satanischer Darstellungen,

  Baden-Baden: Nomos Verlagsgesellschaft.


  GONCZOL, K. (1993) 'The fear of crime' Revue Internationale de Criminalogie 46 (4),



  GRAEF, R. (1993) Living Dangerously: Young Offenders in Their Own Words , London,

  Harper Collins.


  GRANZBERG, C. and STEINBING, J. (eds.) (1980) Television and the Canadian Indians,

  Winnipeg, Manitoba: University of Winnipeg.


  GREENWALD, A. G. (1975) 'Consequences of prejudice against the null hypothesis'

  Psychological Bulletin 82 pp.1-20.


  GREENWOOD, J.(1869) The Seven Curses of London  London: Bew.


  GRIMM, J. (1994) Techniken zur Enbrutalisierung von Medien ms: University of



  GROEBEL, J. (1981) 'Vielseher und Angst' Fernsehen und Bildung, 15, pp. 114-136.


  GUNTER, B. (1985) Dimensions of Television Violence  Aldershot: Gower Press.


  GUNTER, B. (1994) 'The question of media violence' in: J. Bryant and D. Zillman (eds)

  Media Effects: Advances in Theory and Research  Hillsdale, N.J: Lawrence Erlbaum, 



  GUNTER, B. and SVENNEVIG, M. (1987) Behind and in Front of the Screen, London:

  John Libbey.


  GUNTER, B. and  WOBER, M. (1983) 'Television viewing and public trust' British Journal

  of Social Psychology 22, pp. l74-176.


  GUNTER, B. and WOBER, M. (1987) Television and the Fear of Crime London: John



  GUNTER, B. and  WOBER, M.( 1988) Violence on Television: What the Viewers Think 

  London: John Libbey.


  HAGELL, A . and  NEWBURN, T. (1994) Young Offenders and the Media  London:



  HAINES, W.H. (1955) 'Juvenile delinquency and television' Journal of Social Therapy, 1,

  pp. 192-236.


  HALLORAN, J.D., BROWN,R.L. and CHANEY, D.C.(1970) Television and Delinquency 

  Leicester: Leicester University Press.


  HANWAY, J. (1776) Solitude in Imprisonment  London: Bew.


  HARRIS, R. J. (1993) A Cognitive Psychology of Mass Communications  Hillsdale, N.J:

  Lawrence Erlbaum.


  HARTLEY, R.(1964) A Review and Evaluation of Recent Studies on the Impact of

  Violence NewYork: CBS lnc., pp.l9-20.


  HAWKINS, R. and PINGREE, S. (1980) 'Some processes in the cultivation effect'

  Communication Research 7 pp. l93-226.


  HEAROLD, S. (1986) 'A synthesis of 1043 effects of television on social behavior' in: G.

  Comstock (ed) Public Comrnunications and Behaviour: Volume I, pp.65-133. New York:

  Academic Press.



  and CALDER, B J. (1982) 'Impact of the introduction of television crime in the United

  States. Empirical findings and theoretical implications 'Journal of Personality and Social

  Psychology 42,  pp.461-477.


  HICKS, D. (1968) 'Imitation and retention of film-mediated aggressive peer and adult

  models' Journal of  Personality and Social Psychology 2, pp. 97-100.


  HIMMELWEIT, H. T. OPPENHEIM, A. N. and  VINCE, P. (1958) Television and the

  Child: An Empirical Study of  the Effect of  Television on the Young  London: Oxford

  University Press.


  HIRSCH, P. (1980) 'The scary world of the non viewer and other anomalies: Re-analysis of

  Gerbner et al's findings on the cultivation hypothesis' Communication Research 7, pp.403-



  HIRSCH, P. (1981a) 'On not learning from one's own mistakes: A re-analysis of Gerbner et

  al's findings on cultivation analysis.Part 11'  Communications Research 8, pp.3-37.


  HIRSCH, P. (1981b) Distinguishing good speculation from bad theory. Communications 

  Research 8, pp. 73-96.


  HODGE, R. and TRIPP, D. (1986) Children and Television: A Semiotic Approach,

  Cambridge: Polity Press.


  HOME OFFICE (1951) Report of the Departmental Committee on Children and the Cinema 

  London: The Home Office


  HOUGH, M. and  MAYHEW, P. (1985) 'Taking account of crime: Key findings from the

  1984 British Crime Survey' Research Studies No 85, London: HMSO.


  HOUSE OF COMMONS (1994) Video Violence and Young Offenders London: H.M.S.O.


  HOWITT, D. (1972) 'Effects of aggressive modelling upon non-initiative aggression'

  Catalogue of Selected Documents in Psychology 2,  pp.114-15.


  HOWITT, D. (1982) Mass Media and Social Problems Oxford: Pergamon.


  HOWITT, D. and CUMBERBATCH, G. (1972)'Affective feeling for a film character and

  evaluation of an anti-Social act 'British Journal of Social and Clinical Psychology 11, pp.



  HOWITT, D. and CUMBERBATCH, G. (1974) 'Audience perceptions of violent television

  content' Communication Research 1, pp.204-227.


  HOWITT, D. and CUMBERBATCH, G. (1975) Mass Media Violence and Society 

  London: Elek.


  HUESMANN, L. R. (1982) 'Television violence and aggressive behaviour' in: D. Pearl, L.

  Bouthilet and  J. Lazar (eds Television and Behavior Vol 2, Washington DC: NIMH.


  HUESMANN, L. R.  (1986)  'Psychological processes promoting the relation between

  exposure to media violence and aggressive behavior by the viewer' Journal of Social Issues

  42(3), pp. l25-139.


  HUESMANN, L. R and ERON, L. D. (eds) (1986) Television and the Aggressive Child: A

  Cross-National Comparison Hillsdale, New Jersey: Lawrence Elbaum Associates.


  HUESMANN, L. R., ERON, L. D., BERKOWITZ, L. and CHAFFEE (1992) 'The effects of

  television on aggression: A reply to a sceptic' in: P. Suedfield and P.E. Tetlock (eds)

  Psychology and Social Policy New York, NY: Hemisphere , pp.191-200.


  HUESMANN, L. R., ERON, L., KLEEN, L., BRICE, P. and FISHER, P. (1983) 'Mitigating

  the imitation of aggressive behaviours by changing children's attitudes to media violence'

  Journal of Personality and Social Psychology 44, pp. 899-910


  HUESMANN, L. R., KAGERSPETY, K. and ERON, L. D. (1984) 'Intervening variables in

  the TV violence-aggression relation' Developmental Psychology 20, pp.746-775.


  HUESMANN, L. R.and MALAMUTH, N.M. (1986) 'Media violence and antisocial

  behavior' Journal of Social  Issues 42(3).


  HUGHES, M. (1980) 'The fruits of cultivation analysis' Public Opinion Quarterly 44,





  World, Small Screen:  The Role of Television in American Society  Lincoln, Neb:

  University of Nebraska Press.



  effects of TV action and violence on childrens behavior' The Journal of Genetic Psychology

  138, pp. l83-191. 


  JOSEPHS, J. (1993) Hungerford:  One Man's Massacre  London: Batisford.


  JOSEPHSON, W. L. (1987) 'Television violence and children's aggression: testing the

  priming, social script and disinhibition predictions' Journal of Personality and Social

  Psychology 53(5), pp.882-890.


  JOY, L. A., KIMBALL, M. M. and ZABRACK, M. L. (1986) 'Television and aggressive

  behavior' in T. M. Williams (ed.) The Impact of Television: a Natural Experiment Involving

  Three Towns  New York: Academic Press.


  KAY, H. (1972) 'Weakness in the television-causes-aggression analysis by Eron et al'

  American Psychologist 27, pp.970-973.


  KENNY, D.A. (1984) The NBC Study and television violence' Journal of Communication,

  34, pp. 177-188.


  KERN, R. (1986) Medienkonsum, Aggressivit„t und moralisches Urteil bei Jugendlichen -

  Eine Feldstudie ber Bedingungen und Wirkungen des Medienkonsums. Universit„t

  Regensburg: Unver”ffentlichte Diplomarbeit.


  KESSLER, R.C., DOWNEY, G., MILAVSKY, J.R. and STIPP, H.(1988) 'Clustering of

  teenage suicides after television stories about suicides: a re-consideration' The American

  Journal of Psychiatry 145(11), pp. 1379-1383


  KESSLER, R. C. and STIPP, H. (1984) 'The impact of fictional television suicide stories on

  US fatalities: a  replication' American Journal of Sociology 90(1),  pp.l57-167.


  KILLIAS, M. (1983) 'Mass medias: peur du crime et politique criminelle' Revue

  Internationale de Criminologe et de Police Technique 36(4), pp.60-7 1.


  KNIVETON, B. H. and STEPHENSON, G. M. (1970) 'The effect of pre-experience on

  imitation of an aggressive film model ' British Journal of Social and Clinical Psychology 9,



  KREBS, D. (1981) 'Gewaltdarstellungen im fernsehen und die einstellungen zu aggressiven

  hanlungen' Zeitschrift fur: Sozial Psychologies, 12, pp.381-402.


  LAGERSPETZ, K. and VIEMERO, V. (1986) 'Television and aggressive behavior among

  Finnish children' in: L.R. Huesmann and L. Eron (eds) Television and the Aggressive Child,

  Hillsdale, NJ: Lawrence Erlbaum Associates, pp. 81-117.


  LARSEN, O. N. (ed) (1968) Violence and the Mass Media New York: Harper and Row.


  LAZAR, B.A. (1994) 'Why social work should care: Television violence and children' Child

  and Adolescent Social Work Journal, 11, (1), pp. 3-19.


  LAZARUS, R.S. (1962) 'A laboratory study of psychological stress produced by a motion

  picture film' Psychological Monograph 76, whole no. 553.


  LESYNA, K. and PHILLIPS, D.P.(1989) 'Suicide and the media: Research and policy

  implications' in: R.Diekstra The World Health Organization for Preventive Strategies on

  Suicide: A WHO State of the Art Publication.


  LEVEY, L.(1967) 'Awareness, learning and the beneficient subject as expert witness'

  Journal of Personality and Social Psychology 6, pp.365-70.


  LEYENS, J. P., CAMINO, L., PARKE, R. D. and BERKOWlTZ, L. (1975) 'The effect of

  movie violence on aggression in a field setting as a function of group dominance and

  cohesion' Journal of Personality and Social Psychology 32, pp.346-360.


  LEYENS, J. P. and PARKE, R. D. (1975) 'Aggressive slides can induce a weapon's effect'

  European Journal of  Social Psychology 5, pp. 229-236.


  LIEBERT, R. M., NEALE, J. M. and DAVIDSON, E. S. (1973) The Early Window: Effects

  of Television on Children and Youth  Elmsford, New York: Pergamon.


  LIEBERT, R.M. and SPRAFKIN, J.N.(1988) The Early Window: Effects of Television on

  Children and Youth (3rd Edn) New York: Pergamon Press.


  LINNE, O. (1976) 'The viewer's aggression as a function of a variously edited TV film'

  Communications 1, pp.l01-111.


  LISKA, A.E., SANCHIRO, A. and REED, M.D.(1988) 'Fear of crime and constrained

 behavior specifying and estimating a reciprocal effects model'  Social Forces 66(3) pp. 827-



  LISS, M. A., REINHARDT, L. C. and  FREDRIKSEN, S. (1983) 'TV heroes: the impact of

  rhetoric and deeds' Journal of Applied Developmental Psychology 4(2), pp.l75-187.


  LUKESCH, H. (1987) 'Videorecorder und mediennutzung bei Kindern und Jugendlichen

  Rudfunk und Fernsehen, 1, pp. 92-98.


  LYNN, R., HAMPSON, S. and AGAHI, E. (1989) 'Television violence and aggression: a

  genotype - environment correlation and interaction theory'  Social Behaviour and

  Personality 17 (2), pp. 143-164.


  MALETSKE, G. (1959) Television in the Life of Youth  Hamburg: Hans Bredow Institute.


  MAYER, M. (1992) Devianz im Kontext von Medienkonsum und Pers”nlichkeitsfaktoren,

  Regensburg: Unver”ffentlichte Diplomarbeit.


  MAYHEW, P., MAUNG, N.A. and  MIRRLEES-BLACK, C. (1993) The British Crime

  Survey London: H.M.S.O.


  McGUlRE, W. J. (1986) 'The myth of massive media impact: savagings and salvagings'

  Public Communication and Behavior 1, pp. l75-257.


  McHAN, E.J.(1985) 'Imitation of aggression by Lebanese children' Journal of Social

  Psychology l25(5), pp.613-617.


  McKINSEY (1993) Public Service Broadcasting Around the World  cited by P. Golding

  (1994) 'Media research and the new communications map'  The Nordicom Review (1), pp.



  McLEOD, J.M., ATKlN, C.K and CHAFFEE, S.H.(1972) 'Adolescents,parents and

  television use: Self report and other report measures from Maryland and Wisconsin samples'

  in: G. G. Comstock & E. A. Rubenstein TV and Social Behavior, Vol 3 Washington DC:

  US Government Printing Office.


  MclNTYRE, J. J. and TEEVAN, J. J. (1972) 'Television violence and deviant behavior' in:

  G. G. Comstock and E.A. Rubinstein TV and Social Behavior, Vol 3,  Washington DC: US

  Government Printing Office.


  MELTON, A. W. (1962) 'Editorial' Journal of Experimental Psychology 64. pp.553-557.


  MENDELSOHN, H. (1983) 'Using the mass media for crime prevention' (cited by Wober &

  Gunter, 1988).


  MESSNER, S.F.(1986) 'Television violence and violent crime: an aggregate analysis' Social

 Problerns 33(3), pp.218-235.


  MILAVSKY, J R, KESSLER, R. C., STIPP, H. H. and RUBENS, W. S. (1982) Television

  and Aggression: A Panel Study New York: Academic Press.


  MILGRAM, S. and  SHOTLAND, R. L. (1973) Television and Antisocial Behavior: Field

  Experiments New York: Acadernic Press.


  MOLITOR, F. and HIRSCH, K.W. (1994) 'Children's tolerance of real-life aggression after

  exposure to media violence: a replication of the Drabman and Thomas studies' Child Study

  Journal, 24, (3), pp. 191-207.


  MORGAN, M. (1983) 'Symbolic victimization and real-world fear' Human Communication

  Research 9 (2), pp.l46-157.


  MUELLER, C. W., DONNERSTEIN, E. and  HALLAM, J. (1983) 'Violent films and

  prosocial behaviour' Journal  of Personality and Social Psychology 9, pp.83-89.


  MURDOCK, G. (1988) 'Review: The Impact of Television' Current Psychological Research

  and Reviews, 7, (1), pp. 80-82.


  MURDOCK, G. and McCRON, R. (1979) 'The Television and Delinquency Debate' Screen

  Education 30, Spring.


  MUSTONEN, A. and PULKKINEN, L. (1993) 'Aggression in television programs in

  Finland' Aggressive Behavior, 19, (3), pp. 175-183.


  N.l.M.H.(1982) Television and Behavior:Ten years of Scientific Progress and lmplications

  for the Eighties  Rockville, MO: National Institute of Mental Health.


  NATIONAL RESEARCH COUNCIL (1993) Understanding and Preventing Violence,

  Washington, DC: National Academy Press.


  NEWSON, E. (1994) 'Video violence and the protection of children' Journal of Mental

  Health, 3, pp.221-226


  NOBEL, G. (1971) 'Some comments on the nature of delinquents' identification with

  television heroes, fathers and best friends' British Journal of Social and Clinical Psychology,

  10, pp.172-180.


  NOBEL, G. (1975) Children in Front of the Small Screen  London: Constable.


  OLWEUS, D. (1980) 'The consistency issue in personality psychology revisited - with

  special reference to aggression' British Journal of Social and Clinical Psychology 19,




  ORNE, M. T. (1969) 'Demand characteristics and the concept of quasi-experimental

  controls' in: R. Rosenthal and R. Rosnow (eds) Artifact in Behavioral Research  New York:

  Academic Press, pp. 143-179.


  ORNE, M. T. and HOLLAND, C. H. ( 1968) 'On the ecological validity of laboratory

  deceptions ' International Journal of Psychiatry 6, pp.282-293.


  PAGE, M. and SCHEIDT, R. (1971)  'The elusive weapons effect: demand awareness,

  evaluation and slightly sophisticated subjects' Journal of Personality and Social Psychology

  20, pp.304-3 18.


  PAIK, H. and COMSTOCK, G. (1994) 'The effects of television: A meta-analysis' 

  Communications Research 21,  pp.516-546.


  PALMGREEN, P. and RAYBURN, J.D. (1985) 'An expectancy-value approach to media  

  gratifications' in: K. E. Rosengren et al (eds) Media Gratification Research, Beverley Hills,

  Calif: Sage, pp. 61-72.



  (1977) 'Some effects of  violent and non-violent movies on the behavior of juvenile

  delinquents' in: L Berkowitz (ed.) Advances in Experirnental Social Psychology Vol 10, 

  New York: Academic Press.


  PATTERSON, G.R., LITTMAN, R.A. and BRlCKER, W.(1967) 'Assertive behavior in

  children: a step towards a theory of aggression' Monographs of Social Research in Child

  Developrnent 32.


  PAYNE, M.A. (1993) 'Barbadian adults' perceptions of eighteen popular television

  programs' Perceptual and Motor Skills, 77, (3,1) pp. 771-775.


  PEARSON, G. (1983) Hooligan:  A History of Respectable Fears  London: Macmillan.


  PEARSON, G. (1984) 'Falling Standards:  A short sharp history of moral decline' in:

 M.Barker (ed.) The Video Nasties:  Freedom and Censorship in the Media  London: Pluto



  PEPE, A., CROUCH, J. and  EMERSON, M. (1977) 'Violence and television' New Society

  41, pp.536-538.


  PFUHL, E. (1960) The Relationship of Mass Media to Reported Criminal Behaviour 

  Unpublished Doctoral Dissertation, Washington, DC: Washington State University.


  PFUHL, E. (1970) 'Mass media and reported delinquent behavior' in: M. Wolfgang (ed.)

  The Sociology of Crime and Delinquency  New York, NY: Wiley.


  PHILLIPS, D. P. (1982) 'The impact of fictitional television stories on US adult fatalities:

  new evidence on the effect of the mass media on violence' American Journal of Sociology

  87(6), pp.1340-59.


  PHILLIPS, D. P. ( 1983) 'The impact of mass media violence on US homicides' American

  Sociological Review 48, pp.560-68.


  PHILLIPS, D.P. and CASTERSEN, L.L .(1986) 'Clustering of teenage suicides after

  television stories about suicide' The New England Journal of Medicine 315, pp. 685-689.


  PHILLIPS, D.P.and HENSLEY, J.E.(1984) 'When violence is rewarded or punished: the

  impact of mass media stories on homicide' Journal of Communication 34, pp.101-116.


  PHILLIPS, D. P. and PAIGHT, D. J. (1987) 'The impact of televised movies about suicide:

  a replicative study' The New England Journal of Medicine 317, pp.809-811.


  PlLlAVlN, l.A., DOVlDlON, J.F., GAERNTNER, S.L .and CLARK,

  R.D. (1981) Emergency lntervention New York: Academic Press.


  PLATT, S.(1987) 'The aftermath of Angie's overdose:is soap (opera) damaging your health?'

  British Medical Journal  294,pp.954-957.


  PRANK, R. E. & GREENBERG, M. G. (1980) The Public's use of  Television: Who

  Watches and Why London: Sage.


  PRELL, M. (1992) Filmkonsum, moralisches Urteil und Delinquenz, Universit„t gensburg,

  Unver”ffentlichte Diplomarbeit.


  REDFERN, S. (1988) Violence and the Media. London, UK: BBC Publications.


  ROBINS, L. N. (1991) 'Conduct disorder' Journal of Child Psychology and Psychiatry, 32,



  ROBINSON, J.P. and BACHMAN, J.G. (1972) 'Television viewing habits and aggression'

  in: G.A. Comstock and E.A. Rubinstein (eds) Television and Social Behaviour, Vol 3,

  Washington: DC, U.S. Government Printing Office, pp. 372-382.


  ROSENGREN, K. E., WENNER, L. A. and  PALMGREEN, P. (eds) (1985) Media

  Gratifications: Current Perspectives Beverley Hills, California: Sage.


  ROSENGREN, K.E. and WINDAHL, S. (1989) Media Matter: TV Use in childhood and

  adolescence New York, NY: Academic Press.


  ROSSI, P.H. and WRIGHT, J.D. (1983) 'Review and comment: Television and aggression:

  a panel study' Television Quarterly, 6, pp. 41-44.


  ROWLAND, W. D. (1983) The Politics of  TV  Violence: Policy uses of Communications

  Research Beverley Hills, California: Sage.


  RUSHTON, J. (1979) 'Effects of prosocial television and film material on the behavior of

  viewers' in: L. Berkowitz (ed.) Advances in Experirnental Social Psychology Vol 12, pp.

  322-351 New York: Academic Press.


  RUSSELL, C.E.B. (1905) Manchester Boys  Manchester: Manchester University Press.


  SAMPSON, R.J. and LAUB, J.H. (1993) Crime in the making: Pathways and turning points

  through life Cambridge, Mass: Harvard University Press.


  SANSON, A. and DI MUCCIO, C. (1993) 'The influence of aggressive and neutral cartoons

 and toys on the behaviour of school children' Australian Psychologist, 28, (2), pp. 93-99.


  SCHEUNGRAB, M. (1990) 'Die Abbildung von Beziehungen zwishen Medienkonsum und

  Delinquenz im Rahmen kausalanalytischer Modelle'  in: H Lukesch (Hrsg.) "Wenn Gewalt

  zur Unterhaltung wird ..." Beitr„ge zur Nutzung und Wirkung von Gewaltdarstellungen in

  audiovisuellen Medien, Regensburg: S. Roderer, pp. 119-148.


  SCHEUNGRAB, M. (1991) ''Die Abbildung von Beziehungen zwischen Medienkonsum

  und Delinquenz im Rahmen kausalanalytischer Modelle' Archiv fr Psychologie, 142, 4, pp.



  SCHMIDTKE, T. and HAFFNER, W. (1988) 'The Werther effect after television films: new

  evidence for an old hypothesis' Psychological Medicine, 18, pp. 665-676.


  SCHRAMM, W., LYLE, L. and  PARKER, E. B ( 1961 ) Television in the Lives of Our

  Children: Stanford University Press.


  SCHUCK, J and PISO, K. (1974) 'Evaluating an aggression experiment by the use of

  similating subjects' Journal of Personality and Social Psychology 29(2), pp.181-188.


  SHUTTLEWORTH, F. and MAY, M. (1933) The Social Conduct and Attitudes of Movie

  Fans New York: Macmillan.


  SINGER, D.G. and SINGER J. L. (1980) 'Television viewing and aggressive behavior in

  pre-school children: a field study' Annals of the New York Academy of Science, 347,



  SINGER, D. G.,  ZUCKERMAN, D. M. and  SINGER, J. L. (1980) 'Helping elementary

  school children learn about TV' Journal of Communication 30, pp.84-93.


  SINGER, J. L. and SINGER, D. G. (1981) Television, Imagination and Aggression: a Study

  of Pre-schoolers Hillsdale, New Jersey: Lawrence Erlbaum and Associates.


  SINGER, J. L. and SINGER, D. G.(1986) 'Family experience and television viewing as

  predictors of children's imagination, restlessness and aggression' Journal of Social Issues 42,

  (3), pp.l07-124.


  SMITH, D. (1993) The Sleep of Reason  London: Batisford.


  SMITH, D. J. (1995) 'Youth crime and conduct disorders' in: M. Rutter and D. J. Smith

  Psychosocial Disorders in Young People, Time Trends and their Causes  Chichester: John

  Wiley, pp.389-489.


  SMITH, M. L. (1980) 'Publication bias and meta-analysis' Evaluation in Education 4, pp.22-



  SPIVACK, G., MARCUS, I. and SWIFT, M. (1986) 'Early classroom behaviors and later

  misconduct' Developmental Psychology 22(1), pp. 124-131.


  SPRAFKIN, J., GADOW, K. D. and KANT, G. (1988) 'Teaching emotionally disturbed

  children to discriminate reality from fantasy on television' Journal of Special Education

  21(4), pp.99-107.


  SPRAFKIN, J., SWIFT,C. and HESS, R.(1983) Television: Enhancing the Preventative

  lmpact of TV  New York: Haworth Press.


  STEIN, A. H. and FREIDRICH, L. K. (1975a) 'The effects of television content on young

  children' in: A. D. Pick (ed.) Minnesota Symposium on Child Psychology Vol 9.

  Minneapolis: University of Minnesota Press.


  STEIN, A. H. and FREIDRICH, L. K. (1975b) Personal correspondence.


  STIPP, H. (1993) 'The challenge to improve television for children' in G. Berry and J.K.

  Asamen (eds) Children and Television,  Newbury Park, Calif: Sage, pp. 296-302.


  STIPP, H. and MILAVSKY, J.R. (1988) 'U.S. television's programming's effects on

  aggressive behaviour of children and adolescents'  Current Psychology:  Research and

  Reviews 7, pp.80-102.


  STROMAN, C.A. and SETTGER, R.(1985) 'Media use and perceptions of crime'

  Journalism Quarterly 62(2), pp.340-345.


  TAMBORINI, R., ZILLMAN, D.and BRYANT, J. (1984) 'Fear and victimization'

  in:R.H.Bostrum(ed) Communication Yearbook No 8,  Beverly Hills: Sage. pp.492-513.


  TANNENBAUM, P. (1971) 'Studies of Film-and-Television-Mediated Arousal and

  Aggression: A progress report'  in: G. A. Comstock, E. A. Rubenstein and O. P. Murray

  (eds). Television and Social Behavior, Vol 5,  Television Effects  Washington DC: US

  Government Printing Office.


  TANNENBAUM, P.H. and ZILLMANN, D. (1975) 'Emotional arousal in the facilitation of

  aggression through communication'  in: L. Berkowitz (ed.) Advances in Experimental Social

  Psychology, Vol 8, New York, NY: Academic Press, pp. 149-192.


  TAYLOR, L and MULLAN, R. (1986) Uninvited Guests, London: Chatto and Windus.


  THOMSON, R.J. (1959) Television Crime Drama:  Its impact on children and adolescents 

  Melbourne: Cheshire.


  THORNTON, W. and VOIGHT, L. (1984) 'Television and delinquency' Youth and Society

  15 (4), pp. 445-468.


  TURNER, C. W., HESSE, B. W. and  PETERSON-LEWIS, S. (1986) 'Naturalistic studies

  of the long-term effects of television violence' Journal of Social Issues 42(3), pp.51-73.


  TURNER, C. W. and SIMONS, L. S. (1974) 'Effects of subject sophistication and

  evaluation apprehension on aggressive responses to weapons' Journal of Personality and

  Social Psychology 30, pp.341-348.


  TYLER, T. M. and COOK, F. L. (1984) 'The mass media and judgement of risk' Journal of

  Personality and Social Psychology 47, pp.693-709.


  U.N.E.S.C.O. (1961) The Influence of the Cinema on Children and Adolescents  Reports

  and papers on Mass Communications No.31, Paris: U.N.E.S.C.O.


  VAN DER VOORT, T.H.A. and VALKENBURG, P.M. (1994) 'Television's impact on

  fantasy play: a review of research'  Developmental Review, 14, pp. 227-251.


  VIEMERO, V. (1986) Relationship between Filmed Violence and Aggression , Turkin,

  Finland, Reports from the Department of Psychology at Abo Akademi, Monograph

  Supplement 4.


  VIEMERO, V. (1990) 'Results of Finnish studies of the effects of watching TV on the

  viewer' in U. Carlsson Medier Manniskor Samhalle, Nordicom: NYTT, pp.85-92.


  VOGELGESANG, W. (1991) Jugendliche Video-Cliquen. Action und Horrorvideos als

  Kristallisationspunkte einen neuen Fankultur  Opladen: Westdeutscher Verlag.


  VOOIJS, M.W. and VAN DER VOORT, T.H. (1993) 'Learning about television violence:

  the impact of critical viewing curriculum on children's attitudinal judgements of crime

  series' Journal of Research and Development in Education, 26, (3), pp. 133-142.


  WALTERS, R. and THOMAS, E. (1963) 'Enhancement of punitiveness by visual and

  audio-visual displays' Canadian Journal of Psychology, 17, pp. 244-255.


  WARTELLA, E. (1988) 'The public context of debates about television and children'

  Applied Social Psychology Annual, (8), pp.59-68.


  WARTELLA, E. (1995) 'Media and problem behaviours in young people' in M. Rutter and

  D.J. Smith (eds.) Psychological Disorders in Young People: Time Trends and their Causes,

  Chichester: John Wiley, pp. 296-323.


  WARTELLA, E. and REEVES, B. (1985) 'Historical trends in research on children and the

  media' Journal of Communication 35, pp. l l 8-133.


  WATKINS, L. and  WORCESTER, R. M. (1986) Private Opinions Public Polls  London:

  Thames & Hudson. 


  WEIá, R.H. (1990) 'Horror-Gewalt-Video-Konsum bei Jugendlichen.  Gefhlsreaktionen -

  Pers”nlichkeit - Identifikation T„ter/Opfer'  in: H. Lukesch (Hrsg.) "Wenn Gewalt zur

  Unterhaltung wird ... " Beitr„ge zur Nutzung und Wirkung von Gewaltdarstellungen in

  audiovisuellen Medien  Regensburg: S. Roderer pp.47-93.


  WELLS, W. D. (1973) Television and Aggression: a Replication of an Experimental Field

  Study Unpublished manuscript. Graduate School of Business, University of Chicago.


  WERTHAM, F. (1954) The Seduction of the Innocent  New York, N.Y: Rinehart.


  WEST, D.J. and FARRINGTON, D.P. (1977) The Delinquent Way of Life London:



  WESTLAND, G. (1978) Current Crises of Psychology London: Heineman.


  WIEGMAN, O., KUTTSCHREUTER, M. and BARDA, B. (1992) 'A longitudinal study of

  the effects of television viewing on aggressive and pro-social behaviours' British Journal of

  Social Psychology, 31, pp. 147-164.


  WIKSTROM, P. O. H. (1990) 'Age and crime in a Stockholm court' Journal of Quantitive

  Criminology, 6, pp.61-83.


  WILLIAMS, T.M., ZABRACK, M.L. and JOY, L..A. (1982) 'The portrayal of aggression

  on North American Television' Journal of Applied Social Psychology 12, pp.360-380.


  WILSON, B.J., HOFFNER, C. and CANTOR, J. (1987) 'Children's perceptions of the

  effectiveness of techniques to reduce fear from the mass media' Journal of Applied

  Developmental Psychology, 8, pp. 39-52.


  WILSON, P. (1994) Screen Violence and Child Mental Health  London, Young Minds

  (National Association for Child and Family Mental Health).


  WINSTON, B. (1986) Misunderstanding Media  Cambridge, M.A: Harvard University



  WlLLlAMS, T. M. (ed.) (1986) The lmpact of Television: A National Experiment lnvolving

  Three Towns New York: Academic Press.


  WOBER, M. B. and GUNTER, B. (1982) 'Television and personal threat: fact or artifact? A

  British survey'  British Journal of Social Psychology 21, pp.239-247.


  WOBER, M. B. and GUNTER, B. ( 1988) Television and Social Control   Aldershot:



  WOOD, W.,  WONG, F. and CACHERE, J. (1991) 'Effects of media violence on viewers

  aggression in unconstrained social interaction'  Psychological Bulletin, 109,  pp.371-383.


  WORSLEY, H. (1849) Juvenile Depravity  London: Gilpin.


  ZILLMAN, D.(1978) 'Attribution and mis-attribution of excitatory reactions' in:J.H.Harvey,

  W.J.lckes and R.F.Kidd (eds) New Directions in Attribution Research Vol 2. Hillsdale, New

  Jersey: Erlbaum.


  ZILLMAN, D. ( 1979) Hostility and Aggression Hillsdale: New Jersey, Erlbaum.


  ZILLMAN, D. (1982) 'Television viewing and arousal' in: D. Pearl, L. Bouthilet and J.

  Lazar (eds) Television and Behavior, Ten years of scientific progress and implications for

  the eighties Washington DC: US Government Printing Office.


  ZILLMAN, D. (1985) 'The experimental explorations of gratifications from media

  entertainment' in: D. Zillman and J. Bryant (eds) Selective Exposure to Communication,

  Hillsdale, NJ, pp. 225-239.


  ZILLMANN, D. (1991) 'Television viewing and physiological arousal'  in: J. Bryant and D.

  Zillmann (eds) Responding to the Screen: Reception and Reaction Processes, Hillsdale, NJ:

  Lawrence Erlbaum Associates, pp. 103-133.


  ZILLMAN, D. and BRYANT, J. (1986) 'Exploring the entertainment experience' in: J.

  Bryant and D. Zillman (eds) Perspectives on Media Effects  Hillsdale, New Jersey:

  Lawrence Erlbaum, pp.303-324.


  ZILLMAN, D. and  JOHNSON, R. (1973) 'Motivated aggressiveness perpetrated by

  exposure to aggressive  films and reduced by non-aggressive films' Journal of Research in

  Personality 7, pp.261-276.