|
Abstract
|
Two experiments examined the influence of children's cartoons and their prosocial messages on the behavior and judgments of kindergarten, second, and fourth grade children. Three conditions or types of programs were used: purely prosocial (moral lesson but no aggression); and purely aggressive (no moral lesson). Behavioral, evaluative, and comprehension measures were taken. All children were able to distinguish the protagonist and antagonist. Older children grasped the implications of the stories but younger children, especially those in the prosocial/aggressive condition, did not. Scores of children in the prosocial/aggressive condition were significantly lower on message comprehension than those of subjects in the purely prosocial condition. Implications for the relationship of behaviors (prosocial actions) to judgments, portrayal of aggressive heroes on television programs, and the acceptance of aggression as normative behaviors, are discussed.
VioLit summary:
OBJECTIVE:
The purpose of this article by Liss et al. was to examine the influence of the prosocial messages in children's cartoons on the behavior and judgments of kindergarten, second and fourth grade children.
METHODOLOGY:
An experimental design was employed for two experiments. The first experiment studied 60 white middle class kindergarten children (mean age 5.6 years). The children were randomly assigned to one of three groups until each cell contained equal numbers of male and female children. The setting of the study was an elementary school classroom. Three commercial Saturday morning cartoons (about five minutes each) were recorded. These independent measures were labeled purely prosocial (PP) (this portrayed a moral without any aggression shown), prosocial/aggressive (PA) (this portrayed the same moral lesson, but used violence), and purely aggressive (A) (this had aggression and no moral lesson). The PP and PA cartoon were both from Superfriends, and the A program was from Popeye. All programs featured humanoid forms and the PP and PA programs discussed the same "lesson." Each subject was brought to the experimental room individually while the experimenter, a white female graduate student, attended to an errand. The first measure of the dependent variable was a behavioral task (modeled after that derived by Liebert and Baron), and measured the subject's willingness to hurt or help another child. The subject was told that a child in the next room was ready to play a game and that the subject could make the game easier or harder for the other child by pressing either a green button (marked help) or a red button (marked hurt). The behavioral measure was the total amount of time each of the buttons was pressed across 15 trials. A second measure of the dependent variable was character evaluation. The experimenter showed the child an angel/devil scale fashioned after Buchanan and Thompson's spank scale. The scale contained a series of six figures (three angels and three devils in proportionally increasing size). The child was told to circle the figure that best described the hero and the villain. A third measure was gathered during the return trip to the subject's classroom, and measured lesson comprehension. The second experiment examined changes in the behavioral effects and message comprehension across age groups. An experimental design was employed using 120 white middle class children. They were equally represented by males and females at each of three grade levels (K, 2nd, and 4th). Participants were randomly assigned to one of two conditions, prosocial, or prosocial aggressive, creating 12 cells (sex, grade, condition). The procedure was the same as in experiment 1. The post-stimuli interview was expanded to include the evaluation of in-depth understanding of the lesson rather than just a cursory reference. Subjects were asked if anyone transgressed or did heroic deeds. They were also asked if these deeds were done by good guys or bad guys. A coder scored the subjects' answers based on an a priori system determined by the researcher. Points were given for citations of prosocial lessons and discussion of the bad deeds as well as heroic actions. Seven questions which referred to complete answers pertaining to plot, lesson, and reasons for heroic figures' interventions and actions; high scores reflected understanding of lesson and judgments of characters. ANOVA and chi-square were used to analyze the data.
FINDINGS/DISCUSSION:
For experiment 1, a 2 (sex) by 3 (TV condition) by 2 (character type) ANOVA was conducted to test the hypothesis that children can identify protagonists and antagonists. All protagonists were significantly judged to be less "naughty" than all antagonists (p<.01). Characters in both prosocial conditions were less naughty overall than those in the aggressive cartoon. The interaction between character type and TV program was significant (p<.01). Protagonists on the PP an PA programs were significantly "better" than those on the aggressive cartoon. Antagonists on the PA program were judged to be naughtier than those on the aggressive program and the PP program (p<.01). To measure modeling and imitation, a 2 (sex) by three (TV condition) by 2 (response class: help/hurt) ANOVA was done. There was a significant interaction between behavioral response and TV condition. Children in the PP condition significantly helped more than hurt. Children in the PA group significantly hurt more than helped. Children in the PP group helped more than children in the other groups (t=25.58 for PA, and t=32.48 for A). Children in the PP group significantly hurt less than those in the PA group. Simple main effects upon which these comparisons are based were significant for helping behavior and marginal for hurting behavior (p<.05).
In experiment 2 the total scores for the interview data were subjected to a 2 (sex) by 3 (grade) by 2 (TV condition) ANOVA to examine comprehension. Two main effects were significant. Children in the PP condition significantly articulated the program's plot and lesson more completely than did children in the PA condition. Increasing grade level significantly increased understanding of the plot and the moral lesson. More children in the PP group mentioned the program's message than children in the PA group (p<.01). Children in the prosocial condition (62%) were more able to tell the moral lesson than children in the prosocial-aggressive group (95% could not). 92% of the children in the PP group explained why the superheroes were good, while only 45 % of the PA group could do so. A 2 (sex) by 2 (condition) by 3 (grade/age) ANOVA was performed to measure imitation, with repeated measures on the last factor (dependent measure of help/hurt). Overall, subjects helped more than hurt the fictional child (p<.01). Children in the PP group helped more than they hurt (p<.01). They also helped more than the PA group (p<.01). There was support for a significant developmental difference (p<.01) that showed older children (grades 2 and 4) helping more than they hurt. There was a significant interaction between gender and condition with PP girls significantly participating more in the game than males in the PP group and more than girls in the PA group.
AUTHORS' RECOMMENDATIONS:
The authors stated that future work should compare the relative influence of television's well-known superheroes with TV heroes who are average mortals performing the some actions and espousing the same moral lessons. It was also argued that consistency in characters/plots in teaching of prosocial lessons is important.
EVALUATION:
This study begins to get at a better understanding of how young children process what they see in cartoons. This study took preliminary findings from its first experiment and more fully explored them in a second, a process which gives more depth and credibility to the findings related to the degree to which the children were able to understand the prosocial messages in cartoons. The increasing of sample size from 60 to 120 also helps in this regard. However, there are some difficulties with generalizability of both samples. The population from which the children were selected was that of white, middle class children; it would be difficult to draw conclusions about ethnic and/or inner city children living close to the poverty line. Also, particularly in the original experiment, the experiment was quite artificial (when would children push buttons to change the level of play of another?) and somewhat leading to the findings. It could be argued that the process of talking about whether there were moral lessons in the cartoons would be a different processing of the cartoons than if the children had simply watched them. (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 - Program-Film Content
KW - Media Violence Effects
KW - Television Viewing
KW - Television Violence
KW - Child Aggression
KW - Child Behavior
KW - Child Prosocial Behavior
KW - Prosocial Skills
KW - Social Skills Development
KW - Cartoons
KW - Prosocial Television
KW - Aggression Causes
KW - Grade K
KW - Grade 2
KW - Grade 4
KW - Elementary School Student
KW - Social Learning
KW - Late Childhood
KW - Middle Childhood |