|
Abstract
|
This study builds on the research concerning television viewing and aggression by extending the external validity, or generalizability, of the dependent variable. We assess the relationship between self-reported television viewing at 8, 10, and 12 years of age and the subsequent commission of a violent criminal act. This study is based on interview data from 48 males incarcerated for violent crimes and 45 nonincarcerated, nonviolent males matched on age, race, and neighborhood of residence during adolescence. Results show that the extent of a respondent's reported television viewing was not, in and of itself, predictive of violent criminal acts. Instead, it was the interaction of heavy doses of television viewing and exposure to either maternal or paternal abuse that related to violent crime. These findings support the efforts of some recent scholars in their attempts to understand why television has a negative effect on only some viewers. The results are discussed in light of the cognitive formulations of neoassociationism, encoding specificity, and the double-dose effect.
VioLit summary:
OBJECTIVE:
The objective of this study by Heath et al. was to examine the relationship between television viewing and subsequent violent criminal behavior. The results were discussed in the framework of cognitive formulations.
METHODOLOGY:
The authors employed a quasi-experimental cross-sectional design with a non-probability sample of 48 males incarcerated for violent crimes. Four criteria had to be filled for inclusion in the study: the offender had to have been convicted of a crime that involved the threat or the use of force on a victim, the offender had to be between the ages of 18 and 25 in order to insure accurate recall of childhood events, the conviction had to be in the county in which the study was conducted, in order for a matched comparison group to be used, and the inmate had to be male, as there were too few female offenders incarcerated for violent crimes to be of use to the study. The resulting sample consisted of subjects who had been convicted of murder, rape, robbery and/or aggravated assault, and included some cases of incest, kidnapping and burglary. The group was incarcerated in institutions in Hennepin County in Minnesota. In order to increase internal validity, individuals in the criminal sample were precision matched as closely as possible with a comparison group, based upon age, race, sex and neighborhood of residence during adolescence. Posters advertised for interested participants, and 45 individuals were found, who had no criminal history, but who were matched with the subjects on sex, age, sociodemographic characteristics and adolescent neighborhood. A one-to-one match on race was not possible, but relative proportions of racial groups in the two groups were similar. The instrument employed in the study was a retrospective survey with both open- and closed-ended interview questions, as well as paper-and-pencil items. The surveys were primarily the same for both groups, with only slight modifications for the matched comparison group. The survey was aimed at assessing subjects' exposure to television and to violence in the home, the degree of cohesion within the family, school and peer relationships, and prior experience with the correctional system. Television exposure was measured by giving subjects `TV Guide' summaries of television programs that had been shown in one week in the years that the subject was 8, 10 and 12 years of age. Respondents were asked to indicate which shows they had watched all the time and those they had seen some of the time, and also to report which shows and characters were favorites between the ages of 8 and 12 years. In a previous study this method of retrospective reporting of viewing achieved a test-retest reliability of .88. Also included in the survey were six items that measured physical abuse of the respondent by the father and by the mother. A measure of residential instability was also included, as related to family disorganization, to determine if the child had remained with his nuclear family until at least 13 years of age. Subjects' involvement in any recreational clubs and team sports was also measured, to control for time available to watch television. Race was included in this study as previous research had found that white characters were more likely to use justified violence, a behavior which had in turn been found to lead to aggressive responses in viewers. The authors suggested that subjects would be more likely to identify with characters of their own race, in which case race could indirectly affect aggression that stems from television violence. Analysis included examination of frequencies, discriminant analysis and ANOVA.
FINDINGS/DISCUSSION:
By matching the two groups on adolescent residence, the authors also insured that they were matched on family size, family constellation at the time the subject was born, number of moves made by the household, socioeconomic status and crime rate in the neighborhood. Any possible spurious effects of these variables were therefore controlled. The authors found that involvement in recreational clubs and participation in team sports did not significantly correlate with time spent watching television, and no interactions were found between either of these activities and television viewing in predicting violent behavior. No relationship was found between residential instability and viewing habits, and no interaction between stability and viewing was found in regard to violence. Among the inmate sample, 34% reported that their parents had regulated the amount of time they had spent watching television; 54% of the matched group reported having such rules in the household. However, subjects in both groups reported that these rules and regulations had not been enforced - 64% of the inmates whose parents enforced time rules were classified in the high television exposure range, compared with 48% whose parents had no such rules. 50% of the comparison subjects whose parents did restrict viewing time were in the high exposure range, compared with 39% of the group whose parents had no time restriction rules. The authors concluded that variation in amount of time spent watching television was not an artifact of parental rules and supervision. Analysis of television effects included the total number of shows watched all the time, and the total watched sometimes by each respondent at 8, 10 and 12 years of age. Portrayal of intentional harm was used to differentiate between violent and nonviolent programs. Whilst total viewing, violence viewing and nonviolence viewing all entered the equations in the discriminant analyses, none of the predictions of violent behavior was particularly strong. Inmates reported watching significantly more violent programs as adolescents than did their matched counterparts, and similar trends were observed for total viewing and nonviolent viewing as well, although these differences did not reach significance. The correlation between violent and nonviolent television viewing was high, precluding any attempt at separating the unique contributions of each. The discriminant analyses also revealed that both paternal and maternal abuse were associated with adult violent behavior. However, an analysis of variance, which extended these findings, suggested that paternal abuse was only marginally related to later criminal behavior (p<.10) - 56% of the high abuse category were inmates, compared with 47% of those who were victims of lower levels of paternal abuse. This effect was overwhelmed, however, by the interaction among paternal abuse, maternal abuse and total television viewing - those respondents who had been exposed to two of these three forms of violence, or to all three forms, were more likely to subsequently become violent themselves than were those exposed to only one or none of these situations. Exposure to one form of violence in childhood and adolescence did not increase the likelihood of adult criminality, and race did not enter as a main effect or as a part of any interaction effects. Whilst inmates were slightly more inclined to list violent programs as their favorites (58% compared with 56% of the comparison group), this difference was not significant. It was the nonviolent group which was more likely to report having a violent favorite character (53% compared with 44% of the inmate group), and which was more likely to believe that events on television could occur in real life (72% and 70% for the comparison and the inmate group, respectively). No race differences were found in choice of favorite character. The authors concluded that high exposure to television violence was associated with subsequent adult criminality if violence was also present in the home. They explained their findings in terms of three cognitive theories: cognitive neoassociationism, encoding specificity and the double-dose hypothesis. The first theory suggests that children who view television violence within a violent setting associate the violence they view with concepts such as conflict resolution or attaining goals. Violence on television might be associated with violence in the home, as a part of normal life that is an effective tool for achieving one's goals. The encoding specificity theory suggests that if children encode television messages in a hostile family context, a later aggressive situation might trigger the use of those images and behaviors that were seen on television in the hostile context. The double-dose effect suggests that receiving the same message from two sources, that violence is acceptable and normal behavior, increases the effect of each of the two sources. The authors suggested that their findings were consistent with all three of these formulations. They also cautioned that their study could not determine if it was the violence upon television, and not some other depiction of life, that was responsible for subsequent violent behavior. They concluded by warning that the use of serious violent criminals from a high-risk group for criminal behavior precluded generalization.
EVALUATION:
The authors present an interesting and informative examination of the effects of television violence upon adult criminal behavior. However, the small sample size and the use of an extreme sample limit the external validity of the findings. Concerns with accuracy of data are present, as the researchers have had to rely upon respondents' memories of events that occurred many years before. Whilst a more thorough examination of the implications of the findings would have been helpful, the discussion of the measurement of variables was clear, and the use of a matched comparison group helped to increase internal validity. Despite the problem of generalizability, this work should be considered as an important and informative addition to research into violence, and should act as a basis for future research in the field. (CSPV Abstract - Copyright © 1992-2007 by the Center for the Study and Prevention of Violence, Institute of Behavioral Science, Regents of the University of Colorado)
KW - Minnesota
KW - Media Violence Effects
KW - Television Violence
KW - Television Viewing
KW - Crime Causes
KW - Violence Causes
KW - Adult Crime
KW - Adult Inmate
KW - Adult Male
KW - Adult Offender
KW - Adult Violence
KW - Male Crime
KW - Male Inmate
KW - Male Offender
KW - Male Violence
KW - Inmate Studies
KW - Exposure to Violence
KW - Long-Term Effects
KW - Child Development
KW - Child Male
KW - Juvenile Male
KW - Juvenile Development
KW - Youth Development
Language: en |