Over the past 50 years there has been a surge of research by social psychologists regarding the development of mass communication, with over 3500 research studies (Harris 2004) investigating media effects. Whilst others forms of media have been discussed, the growth of research seems to parallel the expansion of one particular form of media; television, as a result the majority of research has focused on the effects of TV.
However critics of such research, such as McQuail (1994) are quick to note that, “the entire study of mass communication is based on the premise that the media has significant effects… et there is little agreement on the nature and extent of these assumed effects. ” In short, whilst there is mounting concern and “moral panic” regarding the media and its effects, studies intended to investigate this relationship have yielded little consistent evidence or have key flaws, which question any conclusions drawn. The majority of concern about media effects relates to; television, children and acts of violence, with the intention of forming some type of relationship between them.
Given this approach, this essay will focus on two examples of experimental research, which have attempted to support this, and focus primarily on the negative impact of the media. First, the classical studies by Bandura (1961) will be outlined, to demonstrate where this belief about the negative effects of the media may have originated, next a critique of Bandura’s work will be offered with reference to methodological and theoretical flaws.
Then more recent studies of media effects conducted by Anderson and Dill (2000) will be outlined, to demonstrate that current anxieties regarding media effects relate to others forms of media and that approaches to media effects research has changed. It will then be argued that whilst Anderson and Dill provide a good theoretical framework with GAAM, and a better approach to the study of media effects, there are key problems with the conclusions that they make, which also extent to the majority of effects research more generally.
These flaws concern the direction which researchers are investigating from, the cultural background which may have contributed to the perceived effects of the media and the theory that media effects is grounded in. In Bandura’s classic experiment (1961) seventy-two 3-5 year olds are exposed to a model and an inflatable doll, “Bobo”. The model either behaves in an aggressive or non-aggressive manner towards the doll, and afterwards the model is rewarded by the experimenter, the child is then given a brief task designed to induce mild frustration and then left alone in the company of Bobo.
Observational measures of the child’s behaviour were taken and results supported a social learning theory. Some specific behaviours were copied (e. g. the mallet was used to attack Bobo and verbal utterances were mimicked, “sock him”) by the children, and it was found that the gender of the model had greatest effect on incidences of aggressive behaviour. However, whilst behaviour very generally may have been more aggressive, some of it was not direct imitation of the violence they had seen (e. g. there was no punching), and the behaviour was restricted to Bobo, and no other toy/ character.
This would suggest that whilst some form of imitation may have occurred, it is clearly restricted to specific circumstances, and shouldn’t necessarily be generalised to other. Given that the results of this study were highly specified and couldn’t be generalised, especially in relation to the effects of television, a second major study was carried out. Banduras’ second experiment attempted to address the previous problems, by having children observe videos of other children as opposed to observing live adult models.
The videos depicted either 2 children playing together, 1 child fighting with another and a) winning all the toys or b) loosing all the toys or, in a third condition; no film at all. In the condition where the aggressive child wins all the toys, the child observing the film, behaved more aggressively and imitated what they saw. However this behaviour was restricted to imitating only what they had seen. The findings from this study seemed to suggest that children were successfully copying the aggressive actions they had just viewed.
For many, Bandura’s findings demonstrated the powerful effect that viewing violent events could have on young children, and as such, Bandura is frequently cited as evidence for the negative effects that viewing violent television can have on children, however is this a valid conclusion to reach? Bandura and colleagues presume that when the children are behaving aggressively that they are imitating the model they have seen because, the task has caused them to be frustrated and this is how they have chosen to respond.
However another explanation for their behaviour is that they are imitating the behaviour not necessarily because they are exhibiting violent tendencies, but because of demand characteristics, the children believe that the purpose of the scenario is to copy the model, this is supported by qualitative accounts where children stated, ” There’s the doll we have to hit mummy. ” This could be due to the lack of ecological validity of the experiment.
Whilst Bandura successfully carried out a controlled experiment examining the effect of a given stimulus (a rarity in media effects research) in an attempt to isolate media effects, the mundane realism of the experiment was sacrificed. As such it is then difficult to generalise these results to how the media affects children in differing circumstances, and the validity of the findings could be challenged, as Bandura presumes their laboratory-observed behaviour would mirror actual real life behaviour. Bandura has also been criticised for the way in which he regards his participants.
Bandura and other effects researchers see children as special cases; presumably more easily influenced than adults, and apply their findings to children only. Gauntlett (1998) points out, that many effects researchers fail to carry out adult control groups, to confirm that their responses would differ to those of children. If these effects were however found to be universal, then these studies would fall short of explaining why all individuals are not violent as surely viewing violence would affect all individuals equally.
Banduras’ study in particular relies on the idea that children are passive viewers of the media who simply watch and imitate what they see. Studies have actually shown that children as young a seven can offer criticisms of the media (Buckingham 1996). This reflects an even larger problem in media effects research more generally; there is a tendency to severely underestimate the recipients, who are frequently regarded as inanimate passive receivers, as opposed to intelligent individuals who can be selective in their interpretation of the media.
More recent experimental research has attempted to examine the effect another form of media; video games. Anderson and Dill (2000) attempted to investigate the effects of violent video games by using both an experimental and correlation design to examine not only the short-term effects (like Bandura) but also additionally the long-term effects of playing violent video games.
Anderson’s approach was a more internationalist one than Bandura, as he viewed violent behaviour not in isolation but in conjunction with personality factors, situational factors and with regard to social learning theories, cognitive and excitation models. By combining these factors Anderson presented the General Affective Aggression model (GAAM), which postulates that exposure to violent video games increases aggressive behaviour in both the short term and the long term.
GAAM was partially tested via two studies carried out by Anderson and Dill. Study one had a correlational design, where Anderson and colleagues collected self report measures, relating to trait aggression, delinquent behaviour, academic achievement and a video game questionnaire. The questionnaire required respondents to state their “top five” games and record the frequency and duration of play, additionally, their ratings of violent content/graphics was recorded.
Results from this study suggested that the majority of the undergraduates preferred games that were violent, and exposure to them was more heavily correlated to delinquency, as was time spent playing. Anderson and Dill concluded, “Violent video game play and an aggressive personality account for a major proportion of aggressive behaviour and non-aggressive delinquency. ” Additionally, they suggest that playing violent video games was a significant predictor of delinquency.
However Anderson was aware of the causal limitations of correlational work and carried out an experimental study to determine causality and short-term effects. In study two high/low irritability undergraduates (based on prior testing) were assigned to play either a violent or non-violent game (predetermined by a pilot study). After several practice sessions they played a second game, a competitive reaction time task with a confederate. On winning the task, participants were able to “punish” their “opponent” by administering a blast of white noise; this was taken as a measure of aggression.
Results demonstrated that those who were highly irritable and played the violent video game behaved more aggressively towards their opponent. Based on the results from both studies, Anderson and Dill concluded that violent video games, created an environment for learning and practising aggression resolutions to situations, playing them had a cognitive effect as it primed for aggressive thoughts in both the short term and the long term, as repetitive play ensured easier access to these established scripts.
This, Anderson believes, causes alterations to personality, which predisposes individuals to engage in more aggressive and delinquent behaviour. Whilst Anderson and Dill’s study is successful in its approach to studying media effects because it attempts to see the influence of violent video games as only part of the cause, as it also considers individual differences, the conclusions which they reach are still debateable, as the results were mixed and moderate, especially in relation to measures of aggressive thoughts, which form the cognitive element of Anderson’s model.
Additionally Anderson and Dills study as well as Bandura’s suffer from more general problems of experimental effects research. Generally effects research suffers because it challenges the issue of violence in society backwards (Gauntlett 1998). Researchers look at the media and then attempt to extend from it and make spurious links. Gauntlett argues that if you wish to study violence in society and get some explanations for it, then begin by looking at incidences of violence, and then try to explain it, by looking at the common characteristics of those who commit violent acts.
Studies which have attempted to provide explanations using this approach found that violent offenders show no specific interest in violent shows/games (Hagell and Newburn 1994), so for many young offenders who are acting violently, the media doesn’t seem to be a strong influence on their behaviour. More general critiques of experimental effects research concern the stimulus they deem as violent, these are frequently fictional scenarios ( e. g. ideo games), whilst ignoring often more violent depictions of real life reflected in the news and factual programmes. If experiments are designed to examine the effects of violent media, than all instances of violence in the media should be investigated not simply the ones that are convenient and those that fall under the category of violent entertainment. Additionally it is important to explain media effects with relation to their origin, whilst the media is generally expanding globally, cross-cultural generalisations would be unwise.
Bandura’s study has to be understood in a historical context, it was carried out in the USA during a period of backlash against contemporary ideas, with an influential conservative movement warning of the dangerous effects of the media, which is still prominent today and reflected in the views of Gerbner (1994); “we are awash in a tide of violent representations unlike any the world has ever seen… drenching every home with graphic scenes of expertly choreographed brutality. This is typical of the views mentioned earlier in this discussion, which regard the media as corrupting force acting upon a passive audience, and fail to consider media audience interactions. By regarding the media as a single external force acting on a porous audience, then relationship between the individual and the media becomes oversimplified. Finally effects research has been criticised because of its theoretical roots.
This essay began with a quote be McQuail who noted that effects research is based on a premise, that the media actually has an effect, but there has been inadequate research as to why this might be the case (Gauntlett 1994). Whilst Bandura suggests that children imitate what they see, why they do so, has not been questioned. Once again study of the media has been simplified to what it does, and how people act, not why it has its affect, whilst studies by Anderson make tentative steps towards explaining this, via alterations in cognitive scripts, evidence for this is moderate.
In conclusion experimental effects research suffers from similar problems as other forms of effects research not mentioned here (longitudinal and content analysis studies) and may be over relied upon by individuals because of its intuitive appeal or for more political reasons or simply because if true if would provide a clear causal explanation for societal problems.
To use an experimental method to study an area as dynamic and complex as the media is ambitious and impossible without consideration of multiple factors, which few experimental studies have done so far. Perhaps what is needed is a different approach to studying media effects as suggested by Gauntlett (1998) which focuses more on influences and perceptions of the media as opposed to effects and behaviour.