Eaton and Dominick set about investigating the ways in which toy products have increasingly become the inspiration for the creation of television cartoons and other programmes, to the point where the phrase 'programme-length commercials' has been used to describe such productions. I have chosen to review this article in the light of recent discoveries from my own research with children, where it has emerged that children are acutely aware of cross-media merchandising techniques and have already formed many product preferences as a result of the approaches made by advertisers in this context.
This study concentrated most specifically on the degree of violence seen in such programmes; the production features, with particular emphasis on the types of music heard, and the race and gender of the characters on-screen, so making the findings highly relevant to my own study. Eaton and Dominick addressed five research questions in their study, as follows:
1. Do toy-linked cartoons use more overt merchandising techniques than non toy-based cartoons?
2. Are toy-linked cartoons more violent than non toy-based cartoons?
3. Are the plots of toy-linked cartoons less complex than their non toy-based counterparts?
4. Do toy-linked shows contain more stereotypes than less commercialised shows?
5. Is there a difference in the number and type of antisocial and prosocial acts shown in toy-linked versus non toy-linked cartoons? (p. 68)
Design and Method
Eaton and Dominick chose to study a purposive sample of TV cartoon shows, with the intention to reflect both qualitative and quantitative aspects of the sources. They refer to the work of Stempel (1989), who argued that a small and carefully selected study may, in some instances, yield more useful information than a random, larger sample. They further justified their study by stressing that the cartoons could not be assumed to be equally important because they differed in the degree to which they are connected to product marketing (p. 68). They claim that a purposive study made it possible to identify any differences between the two types of cartoon under consideration (p.69).
The actual sample consisted of 16 hours of children's cartoon TV programmes, sampled from 1988 network and syndicate shows in the USA. All the cartoons that appeared in the April 1988 issues of TV Guide were listed. They determined whether the cartoons were toy-linked or non toy-linked through a two-part process, as follows:
1. The advertising expenditures of toys connected with cartoon programmes were identified, and those programmes ranking high on this criterion were defined as 'highly commercialised'.
2. Based on Toy and Hobby World's rankings of the retail sales of toy products, the most popular toys (linked with TV programmes) were identified (p. 69)
Having worked this far, Eaton and Dominick ensured that there was equal representation of network and non-network toy-linked shows in the sample. In addition, the cartoons were labelled either 'action adventure' or 'other programmes' to ensure that each toy-linked network and syndicated series had its non-commercialised counterpart (p. 70).
Variables and Definitions
The advertising techniques were measured in a number of ways:
1. The coders recorded any time when the toy tie-in product's name was verbally or visually represented in the programme.
2. The way in which the programme used music as an integral part of that programme and product format was coded, either as 'no music', 'mood enhancing' or 'key part of plot and story'.
3. The coders also noted if the programme aired a promotional advertisement for a future show in the series.
The coders also observed the characters appearing on screen:
1. They counted the number of major, minor and other characters.
2. They noted the species of each character (human, robot, animal, ghost, other).
In this study, programme complexity was seen as an index of the degree of depth and sophistication contained in the cartoon's story-line. The coders measured the variables by:
1. counting the number of sub-plots within a programme;
2. describing the settings in each cartoon;
3. recording the time period portrayed;
4. counting the number of references to the real world.
In terms of measuring the degree of violence, Eaton and Dominick used the four-fold categorisation system designed by Barcus (1983):
1. none (no violence in the cartoon)
2. incidental (violence occurs only accidentally, or consists of acts of nature)
3. subordinate (violent acts occur throughout, but are subordinate to the plot)
4. saturated (programme contains acts of violence intrinsic to the plot)
To identify any racial or sexual stereotyping, the coders:
1. measured the representation of women and minority characters;
2. coded the occupational roles;
3. coded the character attire.
Finally, antisocial acts were classified as: assault, detention, threat, shooting, fighting, verbal aggression, fraud, cheating, theft or lying. Prosocial behaviours included altruism, sympathy, resistance to temptation, explaining feelings and control of aggression.
Coding and Reliability
Three graduate students with training in the procedures and definitions of the analysis were used in this study to establish a satisfactory initial level of reliability. To check final reliability, two independent coders analysed a sub-sample of 12% of the study's cartoons. Intercoder reliability levels reached what Eaton and Dominick described as 'acceptable', at 90%.
Results
No significant differences emerged between network and syndicated cartoons, so Eaton and Dominick presented their findings in terms of toy-linked versus non toy-linked shows. Their findings were discussed under various headings as follows:
Overt merchandising
There were found to be differences in the ways that music was used in the programmes. For example: 44% of the toy-linked shows used music as an integral part of the plot. The most striking difference emerged through an analysis of character species. For example: humans constituted only about half of the characters coded. In 11% of toy-based cartoons, the characters were coded as robots, while only one of the 306 characters in non toy-based cartoons were robots. In contrast, the latter cartoon type boasted 41% of animal characters compared with only 24% of the former (p.71/2). One might guess that robotic toys are more marketable than animal toys, with exposure to such toys during normal viewing hours creating a preference in children.
Violence
44% of the toy-linked shows in this sample were classed as 'saturated' with violence, compared with only 6% of the non toy-linked shows. This is indeed a worrying findings, since one can assume that children will imitate the types of actions they see the characters performing, in order to recreate the cartoon atmosphere, if they actually own the corresponding toy. This might indeed be true, when one considers the popularity of such programmes and toy-lines as the Power Rangers. Eaton and Dominick certainly support the criticism that toy-based shows are more violent in content than their non-commercialised counterparts.
Programme complexity
It emerged that most of the programmes in this sample were basically formulas and simple in construction. The only major difference emerged in terms of setting, since 88% of toy-linked programmes were set in the USA compared with only 31% of non-linked shows. Other than this, the programmes were found to be more alike than different (p. 72).
Stereotyping and role portrayals
Generally, there was little difference between the cartoons, since males dominated and accounted for more than 75% of all characters in each cartoon type, and few portrayed members of the minority groups. The dominance of male characters seems to correspond with the findings from other studies dealing with many aspects of the media.
Antisocial behaviour
Product related cartoons were found to have 20% more antisocial acts than non toy-based cartoons (p. 73). Interestingly, though not surprisingly, males were more likely (80%) than females to commit an antisocial act and this did not differ by programme category. This is certainly in-keeping with my own research into children's toy advertisements, since the advertisements aimed at boys are consistently more aggressive in address than those aimed at any other audience category.
Prosocial acts
Non toy-based shows were found to contain slightly more prosocial acts than toy-connected cartoons, and these acts were usually performed by adults (especially women) and children (p.74).
Discussion
Eaton and Dominick admit that the content analysis is subject to limitation, and believe that a subsequent analysis of a larger sample would be valuable in determining that generalisability of this study's results. They also claim that their results may be influenced by the fact that the 'action adventure' genre of toy-linked cartoons does not have an exact counterpart in non toy-linked cartoons. Finally, they state that the next stage in such a study should be to investigate the audience response among children, which is an issue that I hope to deal with in my own research.