TELEVISION VIOLENCE - A REVIEW, SUGGESTED RESEARCH DESIGN AND STATISTICAL APPROACH

The bulk of research investigating the effects of television violence on viewers has concluded that it leads to aggressive behaviour: Whilst the weight of evidence does demonstrate an association between televi sion violence and aggressive behaviour; closer scrutiny of some influe ntial and prominent studies in the area indicates that in a number of cases there exist methodological and/or statistical weaknesses that su ggest a need for further work in the area, particularly if claims of c ausation are to be more confidently advanced. This paper examines some of the methodological limitations that appear in the diversity of res earch designs employed in the literature. Arising from this examinatio n is the proposition of a longitudinal model which, it is hypothesised , overcomes a number of the identified limitations and utilises relati vely recent advances in statistical techniques.