Does television violence enhance program popularity?

VioLit summary: OBJECTIVE: The aim of this study by Diener and DeFour was to examine the effects of fictional television violence upon adventure program popularity. METHODOLOGY: The authors conducted two cross-sectional studies for their research. The first study, quasi-experimental in nature, involved collecting data about the frequency of occurrence of 78 specific events in 71 episodes of twelve adventure programs, which were then coded by multiple coders into categories pertaining to violence, verbal aggression, suspense, emotion, romance, humor and fast action. The items were grouped into four major categories: verbal and physical aggression, drama (including suspense, emotion and romance), action, and humor. Interrater reliabilities were adequate for three of the four categories, but were somewhat low for the drama category. After viewing each episode, the coders rated it on 14 subjective scales, such as perceptions of excitement and of violence. They were also asked if they had liked the episode that they had just seen, and then each completed the California Psychological Inventory. Nielsen popularity index scores were also obtained for each episode, to determine the average audience as measured by the percent of households owning televisions in the United States that were tuned to a program during an average minute of the show. Analysis included examination of frequencies and means, as well as multiple regression, correlations and principal components analysis. The second study was experimental in nature, and involved a non-probability sample of 48 females and 52 males who participated in the project as part of an introductory psychology class. All subjects initially answered questions about their liking of adventure programs and in particular. After this, every other group of students was assigned to a violent version of the show Police Woman with the remainder viewing an edited, nonviolent version of the same episode. The subjects then completed the Personality Research Form, to assess personality factors that could be related to program preference. Analysis included MANOVA and correlations. FINDINGS/DISCUSSION: For the first study, the authors found that the average viewing audience for all 11 adventure shows combined was 21%, with the mean percentage of viewing households tuned to adventure programs was 33%. They concluded that, whilst adventure programs were apparently popular, their popularity was not overwhelming. Turning to results of a multiple regression with program popularity (Nielsen rating) as the dependent variable and scores on the four categories as the independent variables, the authors found that objective ratings were not significant predictors of popularity. Correlations between the four objective rating scores and the Nielsen rating were also small and non-significant, suggesting that qualitative aspects of shows are important to popularity, rather than mere frequency of specific events. Turning to subjective ratings of shows, the authors took the mean for each episode of ratings of all 30 coders. Principal components analysis found three factors: emotional involvement, including the scales of believable, suspenseful, emotional, exciting, dramatic, entertaining and overall liking; violence/action, with scales of violence and action; and non-humorous, with scales of shocking and humorous. The correlation between subjective liking and objective violence content was a low -0.04. The authors suggested that coders' rated liking of a program was not increased when violent content was included. They concluded that factors such as plot and character development might be more likely to increase adventure program popularity than was violence, and that there was no direct causal relation between violence and popularity. Results of the second study showed that, whilst the violent version of the episode was slightly preferred, this difference was not significant. The violence condition accounted for 1.3% of the variance in overall liking of the show. The overall liking score was correlated with the perceived violence score, and for all subjects this correlation was nonsignificant. However, for violent subjects, the correlation was significant and negative, indicating that individuals in the violence group perceiving the program as being more violent liked the program less. Dividing the subjects into four groups based upon sex and condition, and examining personality characteristics of the subjects, the authors found that the Aggression subscale gave significant differences on its correlation with overall liking for the four groups. For subjects in the violence condition, a high Aggression score was positively related to liking for males, and negatively related for females. In the nonviolent group, high Aggression scores were negatively related to liking for males, and positively for females. The difference between males and females in the violent condition, and between males in the two conditions, were both significant, suggesting that aggressive males tended to like violent content and to dislike nonviolent content. The authors concluded that violence in and of itself did not enhance program popularity, and accounted for a trivial proportion of the variance in liking and in popularity in the two studies. AUTHORS' RECOMMENDATIONS: The authors suggested that a broader subject population could be used in future studies, as well as a larger number of programs and programs of other types. Research should also examine other types of violence, and should include the development of a theory of program liking based upon personality needs and transient situationally-induced needs. Multidimensional scaling could determine viewers' categorization of programs, which could then be related to popularity and liking. The relationship between liking of violence and program effect upon behavior should also be explored. EVALUATION: The authors present an interesting examination of the relationship between program popularity and liking and violent program content. Whilst the use of subjective coding in the first experiment raises concerns for generalizability, the use of an experimental design in the second study allows for more faith in the external validity of the results. However, it is difficult to determine how reliable the results are due to a lack of thorough explanation of the measurement and operationalization of variables, and due to a lack of detailed examination of findings. Whilst a more thorough discussion of the implications of the findings, and of recommendations for future research, would have been helpful, overall the studies represent a first, brief step in the study of the relationship between violent content and program popularity. (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 - Media Violence Effects KW - Television Viewing KW - Television Violence KW - Program-Film Popularity KW - Adult Perceptions KW - Program-Film Content KW - Exposure to Violence