Chapter 4

Content Analysis of Children’s Televised Toy Advertisements

Abstract

The main aim of this chapter is to clearly identify the ‘formal features’ used in a sample of children’s televised toy advertisements and to formulate a framework that will become the main point of reference throughout the investigation as a whole (cf. Research Issue 2). The premise of the study is that the formal features used in toy advertisements have distinct gender connotations, so a series of content-related hypotheses were generated based on previous content studies for both adult- and child-advertisements. Content analysis was considered the most effective way to code and count all textual elements because the methodology facilitates a detailed study of the less obvious features within a text. In the context of this study, the methodology identified those (gendered) elements that may have a subtle (albeit perhaps profound) influence upon the audience.

A sample of 117 toy advertisements was accumulated during the 1996 Christmas period. Each advertisement was classified according to the most likely target audience (‘boy’, ‘girl’ or ‘mixed’), judged in terms of product type and the sex of any on-screen characters. The investigator and ten independent adult coders classified the most likely target audiences and the advertisement sample was summarised in terms of ‘tokens’ (the total number of advertisements recorded within a given period) and ‘types’ (the number of different advertisements).

The content features were coded in terms of production and camerawork features (shot angle, shot size, camera movement and lens movement), post-production and editing features (transitions, shot duration and voiceovers), and setting and music. Each content category was defined in accordance with industry terminology and illustrated with either stills from the advertisement sample or sketches/diagrams. The content counts were superficially commented upon before undergoing more rigorous comparisons using the Chi-square test of independence. The emergent content patterns were rather telling in terms of the different production techniques and values for the products targeted at the specific audience sectors. A number of ‘marked’ and ‘unmarked’ categories were formally identified before being applied to the initial set of hypotheses, and the overall findings seemed to support existing research in the field.

4.1 Aims and methodology

Content analysis was chosen for a number of reasons, as outlined previously. Since my intention was to construct a detailed framework of toy advertisements before speculating about how the intended audiences might respond, the method seemed a most effective way to reveal the so-called ‘hidden meaning’ of the commercial texts (Hart, 1991: 108). Indeed, as Leiss et al. (1990: 218) stressed, the counting of content items provides a sound framework or foundation for later studies of interpretation

Having previously considered the method and findings of other content analyses studies, an organised structure needed to be imposed on my own research as a means of outlining and investigating any possible gendered patterns in the sample of toy advertisements. A series of hypotheses was established in an attempt to consider all the advertisement elements. The following statements were eventually formulated:

1.       There are more variable shots in advertisements directed at boys than in those directed at girls i.e. girls’ advertisements employ fewer variations in camera technique (cf. Welch et al., 1979).

2.       There are more fades and dissolves in advertisements aimed at girls (cf. Smith & Bennet, 1990: 99; Welch et al., 1979: 205-7; Verna, 1975, in Courtney & Whipple, 1983: 22).

3.       More male voiceovers than female voiceovers are used overall, even in advertisements promoting ‘female’ products (cf. Fowles, 1996: 208-9; Smith, 1994: 331; Smith & Bennet, 1990: 102; Livingstone & Green, 1986; Manstead & McCulloch, 1981; Welch et al., 1979: 207; Barcus, 1977: 29; Dominick & Rauch, 1971, in Courtney & Whipple, 1983: 17).

4.       Boys’ advertisements contain more shots than any other advertisement type (cf. Smith, 1994: 332; Welch et al., 1979: 205-7).

5.       The average duration of each shot is shorter in advertisements aimed at boys (cf. Welch et al., 1979: 205-7).

6.       Girls’ advertisements are always set in the home environment (cf. Smith, 1994: 333; Peirce, 1989, and Dominick & Rauch, 1971, both in Courtney & Whipple, 1983: 17).

7.       Auditory features, such as sound effects and music, are louder more obvious and more variable in advertisements aimed at boys (cf. Welch et al., 1979: 207).

8.       There are generally more advertisements aimed at boys than at girls (cf. Smith, 1994: 331; McArthur & Eisen, 1976 and Barcus, 1975, both in Courtney & Whipple, 1983: 22; Winick et al., 1973: 27).

These hypotheses will be reassessed in the light of the content analysis findings. Firstly, however, a detailed account of the toy advertisements in this sample will be given, including the coding procedures for the intended target audiences and the manifest content in each advertisement. The results of the content analysis will also be tested using the Chi-Square test (discussed later) and the implications for viewers in terms of possible gender connotations will also be addressed.

4.2 Sampling and coding

Initially, a number of child-directed advertisements had to be collected and it was decided that the sample would focus exclusively on toy advertisements shown in the period leading up to Christmas 1996. Advertisements for breakfast cereals, snack foods, music and videos were excluded since these have all been considered in many other studies (cf. Winick et al., 1973). Advertisements were recorded between 7.00 and 9.30 a.m. every Saturday from November to the mid-December 1996. Saturday mornings were targeted both for reasons of convenience and because there is often a concentration of child-directed advertisements at this time (Condry, 1989: 203). Also, the period leading up to Christmas is characterised by a concentration of commercials promoting toys (ibid..: 188-9; Barcus, 1977: 120). While this may not be typical of the situation at other times of the year it facilitates the accumulation of many toy advertisement examples. The eventual sample consisted of 117 different advertisements.

The next stage was to classify the advertisements according to the most likely target audiences. The toy advertisements therefore needed to be categorised in terms of whether they were intended for boys, girls or a mixed audience. I completed the initial categorisations myself by considering the nature of the actual product being advertised and the sex of the key characters shown on the screen (cf. Bretl & Cantor, 1988). This initial classification was achieved with ease, arguably because it was based on how the advertising companies intended the advertisements to be perceived, where various assumptions are made about the toys that boys and girls are deemed to ‘like’. Where boys and girls were shown interacting with both one another and the product, such as family board games, the advertisement was classified as appealing to both sexes (mixed).

However, in an investigation of this kind there is often an intrinsic problem of bias where an investigator may categorise (media) texts in accordance with the overall research objectives. In this instance, I may have been looking for the features that I considered to be representative of certain audience sectors, observing patterns not intended by the advertisement makers. Therefore, in order to achieve more objective target audience judgements and to ensure an acceptable level of inter-coder reliability, I decided it would be appropriate to take the advertisement sample to a wider audience for further classification.

Since the advertisements were child-oriented it was felt that the coders should be parents to ensure at least some awareness of the toy products available, as well as the product preferences of their own children. One could also argue that the toy advertisements might ‘communicate’ with parents since they invariably make the purchase decisions. Parents may also be able to judge the effectiveness of the advertisement appeals as well as whether or not the on-screen child characters exhibited behaviour patterns similar to their own children. Ten individuals, five males and five females aged between 21 and 55, were selected for the coding task. These individuals were approached because they were parents known to the investigator, facilitating access into their homes and more relaxed coding situations.

Each coder was interviewed in the home environment and shown a 35-minute edited video containing each of the 117 toy advertisements in the sample. The coders were encouraged to consider the most likely target audience in each instance and give oral responses about their categorisations. Perhaps rather predictably, they seemed to base their target audience judgements on the gender of the characters appearing on screen – the presence or absence of boys or girls. This content marker had interestingly also been used by children as a way of identifying target gender (Wright & Huston, 1981: 85). The responses were then compared and the eventual target audience judgement for each advertisement was taken from the ‘majority response’ where six or more coders were in agreement. Following this guide, there was target audience agreement in 115 out of 117 advertisements, resulting in a convincingly high level of inter-coder consensus (98.3%).

Some further patterns also emerged. No advertisement that was judged (by the majority) to be aimed at boys was classified by any coder as being aimed at girls, while no advertisement judged (by the majority) to be aimed at girls was classified by any coder as being aimed at boys. One can therefore surmise that the ‘boy’ and ‘girl’ appeals appear to be very distinct from one another and very much restricted to the realm of ‘masculine’ and ‘feminine’. In instances of uncertainty about the target audience of the advertisement, the coders consistently categorised the text as appealing to a ‘mixed’ audience. One might argue that classifying a border-line advertisement as appealing to a mixed audience is simply a way of ‘sitting on the fence’ or ‘playing safe’ and not confining the appeal of the product to a single target audience. One could therefore argue that the target audience judgements were consistent for each advertisement in the sample. A full summary of the way in which the coders classified the advertisements is given in Appendix B.1 followed by a summary of the way the products spanned the three audience-categorises (B.2).

When summarising the target audience classifications made by the coders, it was effective to adopt the (linguistic-based) distinction between ‘tokens’ and ‘types’. In a written text the total word count is a count of the number of ‘tokens’. A count of the number of ‘types’ would be the number of different words. In the case of my sample of televised toy advertisements, the count of ‘tokens’ was the total number of advertisements, regardless of whether any examples are repeated. A count of ‘types’ was the number of different advertisements shown. For the purpose of this content analysis, I was only interested in the number of advertisement ‘types’. From the inter-coder reliability study of the 117 advertisements, 43 were classed as being aimed at boys, 43 as being aimed at girls and 31 for a mixed audience. A summary of the counts of ‘tokens’ (during the recording period as a whole) and ‘types’ is shown in the table below:

4.2.1: Summary of the ‘tokens’ and ‘types’ in the toy advertisement sample

It is interesting to note that while there were equal ‘types’ of boy- and girl-targeted advertisements there were many more male ‘tokens’. That is to say, those advertisements deemed to be aimed at boys were repeated more frequently than those aimed at girls. It is interesting to consider the possible impact of such a pattern in the context of a ‘typical’ Saturday morning viewing session, with commercial breaks occurring approximately every 20 minutes. It is likely that the viewer would be exposed to more ‘male’ targeted advertising, giving a distinctly ‘male dominated’ feel to the content of each commercial break. Similar patterns emerged in previous research and have been supported by Smith (1994) who noted that there were more single-sex advertisements positioned towards boys than towards girls (55 for boys compared to 27 for girls). Although her sample may have been small, Smith’s findings seem to be consistent with my own and, one could argue, indicative of the world of (children’s) advertising as a whole.

As well as categorising the target audience for each advertisement, the coders were also asked to offer reasons for their decisions. These reasons seemed to be as consistent as their target audience judgements since the same descriptions emerged time and again. Advertisements were categorised as being aimed at girls, for example, because they showed toys that were pink, soft and cuddly, often designed to be played with indoors, and generally marketed to foster mother-baby role-play, household activities and an interest in outer appearance (clothes, jewellery, hair and make-up). In sharp contrast, the advertisements were categorised as being aimed at boys because they showed acts of aggression, used darker colours, emphasised competitive behaviour, and were accompanied by noise and rapid activities (cf. Chapter Seven).

It became clear that the coder classifications were primarily driven by the overt content of the advertisements. The coders were certainly sure that the advertisements aimed at boys and girls respectively were significantly and intrinsically different. The aim of conducting a content analysis of the advertisements seemed to be the best way of moving beyond the overt content of the media texts and to consider more closely the less obvious ways in which they differed. One could argue that the differences in so-called form of the advertisements – the way in which formal production features are utilised – might have a subtle yet profound effect on the target audience. Since the production features are less obvious to the viewer than the narrative structure or theme, it is important and necessary to develop a formal framework of content features in which any differences might be identified. It may even be possible to judge whether some production features are used to target specific genders and whether these subtle features may contribute to the overall ‘gendered appeal’ of many of the toy advertisements.

4.3 Deconstructing the content of the toy advertisements

Once the apparent target audience for each product had been agreed upon, the process of content analysis began in earnest. Since the main aim of this analysis was to develop a clear content structure for each advertisement ‘type’ the various ‘technical’ features within each advertisement needed to be clearly defined, identified and counted. The advertisements were closely scrutinised by taking a shot-by-shot approach, freezing the videotape in each new scene to record all of the various and subtle camerawork features present. A shot may be defined as one uninterrupted view from the camera lens and each time this view changes, either in terms of a definite cut or a dissolve, another shot is counted.

During this painstaking exercise, particular attention was paid to the so-called editing and post-production features identified by Welch et al. (1979) such as transitions, shot duration and voiceovers. In this way, I was able to consider the patterns identified in their small scale study, twenty years ago in the USA, in terms of those seen in my own (significantly larger) sample of contemporary toy advertisements. I also investigated the previously neglected area of production or camerawork features, such as shot sizes and angles, and camera and lens movement. Further notes were also made about the setting and soundtrack in each advertisement.

The general fields of production and post-production techniques were composed of more detailed sub-categories. In order to ensure that my content analysis remained consistent in reference throughout, it was necessary to use particular category and sub-category headings when noting the features of each advertisement. Recognised camera technique classifications were used (Millerson, 1985: 38). Millerson (ibid.) listed the possible ways in which the camera frames a person, from a very close-up face shot to an extra-long full-body shot. He classified each person-shot in terms of ‘the amount of person appearing in the picture’. Millerson’s classifications were too detailed for this investigation so they were simplified for ease of reference. Watts (1984: 130) and Millerson (1985: 49) were referred to for general camera movement classifications. The overall structure of the features recorded in my content analysis is summarised below:

4.3.1: Summary of the content features noted in this analysis

4.4 Defining production and camerawork sub-categories

4.4.1 Shot angles

High, level and low angle shots were the terms used to note the angle at which the camera was held when viewing a particular scene, product or person. A level shot was used to refer to the way that a given scene is viewed at the same level, as if one were looking directly ahead at something within one’s field of vision. A high angle shot was used to describe a situation in which the centre of focus was looked down upon from an elevated position. A low angle shot, on the other hand, was used to describe a situation whereby the centre of focus was looked up to from a diminished position. Additionally, the term overhead was used to describe a shot angle that was too extreme to be coded as high angle shot, being directly over a scene and giving the viewer a so-called ‘bird’s eye view’. Finally, skew or cant was used to refer to a situation in which a given shot was taken from the ‘normal’ position and placed at an angle of approximately 45 degrees within the field of vision, creating a scenario in which the shot appeared ‘dislodged’ to the viewer. This type of shot invariably required the viewer to tilt his/her head in order to see the scene properly.

One further shot technique found in the sample of toy advertisements is the effect termed ‘split screen’. This technique creates the impression that there are multiple television screens in one and offers the viewer a number of different perspectives simultaneously. The image below illustrates how a ‘split screen’ appeared in an advertisement for an Action Man product. The left of the screen shows film footage of a man acting out the part of Action Man, while the right of the screen shows the actual product:

4.4.2 Shot sizes

Only three shot-size references (long, mid and close-up) were used in order to maintain a degree of simplicity in an already complex analysis. Long shot was used to refer to all product stills at the end of each advertisement, as well as ‘full body’ shots of any individuals. A ‘full body’ shot, in this instance, was any shot in which the whole body of an individual, or the head, torso and legs-to-the-knee are shown. Mid shot was used to refer to the display of a product too close to be regarded as a long shot or to a shot of an individual from the waist up. Close-up was used to refer to any shot of product detail or an above-shoulders view of an individual. For ease of reference, this term was also used to include ‘extreme close-ups’ such as face shots.

To ensure that the eventual structure of the content analysis was explicit I decided to further divide the shot size counts into two sub-categories. I counted the size of the shots framing all human elements in the advertisements, including male and female adults and children (‘character shots’). I also counted the shots framing all the non-human elements, including the actual products, any puppets and all settings/establishing shots (‘product shots’). This sub-categorisation of the shots would also identify the degree to which one sex may be presented differently from another or any differences in the presentation of the actual products according to audience category.

4.4.3 Camera and lens movements

Pan right and pan left described the way in which the camera ‘looked across’ a scene from a static base-position, to capture a panoramic feel. Perhaps this movement best mimics the movement of a human head, looking from one side of a tennis court to another as each shot is played. Ped (or pedestal) up and ped down were terms used to describe the way in which the camera can be moved up and down its own ‘spine’. That is to say, the camera moves vertically up and down in the same base-position, showing a scene from different level angles. Tilt up and tilt down were distinct from peds in that the nose of the camera moved independently to produce different shot angles. This technique basically involved a ‘nodding’ movement, capable of depicting a smooth movement from a high to a low angle. Zoom-in and zoom-out shots were self-explanatory in that they refer to the way in which a camera may close in on the given centre of focus to give a closer view, or open out from the centre of focus to provide a shot of the surrounding scene. Zoom-in and zoom-out techniques often occurred at varying speeds.

In the majority of the advertisements, the camera focus was sharp. Since this was the ‘norm’, one could argue that it would be unnecessary to note it as a significant technical feature. The only notable focus issue was that of blurring, referring to instances in which a scene was out-of-focus or in soft-focus. The term blurring was, in this instance, borrowed from photography, referring to the way in which movement is captured by overlong exposure (cf. Spitzing, 1974: 42-6). Finally, slow motion and high speed were used to refer to the speed at which a particular piece of film footage was played back. The former may be used to provide greater clarity and detail for the object in focus in the same way as ‘action replays’ during televised sports’ coverage, while the latter is often used to blur and suggest rapid movement.

4.5 Defining post-production and editing sub-categories

4.5.1 Transitions

Cuts, dissolves and fades are collectively referred to as ‘transitions’. A cut describes the clean break dividing one shot from another. Dissolves and fades are used to mean the same in this study, and refer to the way in which two frames of film are laid on top of one another; while one frame fades out the other is dissolved in so that no clear cut is visible. This three-stage process can by clearly seen by capturing stills before, during and after the occurrence of a dissolve as it appears in a television advertisement. A cut is generally regarded as a harsh transition while dissolves and fades are classed as soft or gentle transitions. Another transition noted is one that I have labelled the ‘swirl cut’, a form of wipe (cf. Millerson: 1985). It refers to a static shot of a particular scene that is seen to ‘swirl’ away from or towards the viewer before a second static shot appears on the screen.

4.6 Testing the data

Noting even these few features generated an enormous amount of data since each shot in the advertisement sample often comprised a combination of many techniques. One can begin to appreciate the extent of the detail involved when one notes the total number of shots in the sample as a whole. The 43 boy-targeted advertisements, for example, contained a total of 740 shots, while the same number of girl-targeted advertisements contained 506 shots with a further 458 shots in the 31 mixed audience advertisements. Therefore, the counts featured in this content analysis are based on a total of 1,704 carefully composed and analysed shots. A detailed breakdown of all the shot compositions in this sample is provided in the Appendices C-E, which is broken down into a clear summary of the counts for each advertisement (Appendix B.6). For ease of reference, a clear summary of count distribution across all the coding categories is given in Appendix B.3.

Before applying any formal methods of analysis to the content data it was possible to observe a number of distinctive patterns. For example, the most frequently used shot angle was the level angle, while there were a greater number of low angle shots in advertisements aimed at boys. Boys appeared more frequently in girl-targeted advertisements than girls did in boy-targeted advertisements. Indeed, the appearance of girls in boy-targeted advertisements was extremely rare. In addition, mixed advertisements used more shots of boys than of girls, perhaps indicating that they were predominantly targeted at a male audience even though the product was arguably suitable for both sexes. Another telling pattern emerged in terms of the transitions used. It was clear that boys’ advertisements used a significant number of cuts, while the girls’ advertisements employed more fades and dissolves.

Having made some initial observations, I set about making more exhaustive comparisons of the various camera techniques. To determine whether the obtained data were significantly different, I applied the Chi-Square statistical test to the variables I wished to compare. The main purpose of using the Chi-Square test was to be able to make comparisons between data counts by calculating the degree to which two variables were independent of one another. This was deemed particularly suited to this advertisement data because the same (content) variables had different counts within each audience sector, whilst there were equal numbers of ‘types’ for the key target audiences (43 boys’ ads and 43 girls’ ads), making any comparisons logical and balanced.

Hence, it was hoped that the Chi-square test would effectively formalise and clarify the (sometimes subtle and unobtrusive) degree to which the sample advertisements were ‘different’ in their use of various content features. Any difference was then described as being significant, highly significant or very highly significant. The test also eliminated the possibility of any ‘chance’ relationships emerging. The test was also appropriate in this study because it is designed to compare two or more independent samples, such as boys’, girls’ and mixed audience advertisements respectively, where the data counts are arranged in categories, such as long, mid and close-up shots (Wimmer & Dominick, 1991: 234 ff.). It must be emphasised, however, that the test provided little information about the strength or form of the association between the variables. Furthermore, even if the testing of some variables proved invalid using the Chi-Square test, it did not follow that the relationship between the variables was insignificant. For example, the test will often reject comparisons including zero values, though any contrasts between a zero-value and a large number will often be highly significant and important.

For the test to be a useful indicator of significant difference the variables being tested must be reasonably comparable. For example, comparing the use of high, low and level-angle shots would be a productive and telling exercise, but there would be very little point comparing two unrelated sets of variables such as zooms and transitions. I systematically worked my way through the list of content counts, comparing as many variables as possible. It soon became apparent that a number of interesting patterns existed within the data sets that were often not obvious through simple observation. The Chi-Square results for the variables within both the production and post-production content counts will now be considered, highlighting those differences that emerged as statistically significant.
 

4.7 Testing production and camerawork variables

4.7.1 Shot angles

One could argue that the choice of shot angle in a given advertisement scene is particularly significant for the viewer. The viewer may either be placed as a detached purveyor of the scene by a level shot, for example, or actually be placed as a participant in the action by a high angle shot, for example. Each shot choice may convey a different ‘feel’ in an advertisement, so a study of the differences in shot use across the audience categories is an important consideration. For ease of reference, the various shot angle counts for each audience category in the advertisement sample is summarised in the table below [4.7.1.1]:

4.7.1.1: Summary of shot angles

The Chi-Square test was run on all comparable sets of data in this instance. Perhaps the most important differences emerged when comparing the use of level, high and low angled shots across the target audience categorises. The test revealed an extremely significant difference in shot angle (p<0.0001) when comparing boys and girls. Additionally, a very highly significant difference (p<0.001) emerged when girls’ advertisements were compared with mixed advertisements. A significant difference (p<0.05) also emerged when comparing boys’ and mixed audience advertisements, although the level of significance was not as high as in the other comparisons. The Chi-Square [c² ] results are summarised below [4.7.1.2]:

4.7.1.2: Comparing three shot angles

Since the most frequent angle was the level shot, I eliminated the variable from the next set of tests. A significant difference (p<0.05) emerged when comparing boys’ and girls’ advertisements, and also when comparing boys’ and mixed advertisements. Interestingly, there was no significant difference when testing the high and low angles used in girls’ and mixed advertisements so one could argue that, in this instance, the advertisements aimed at boys made the most noteworthy use of the two extreme angles. These results are summarised below [4.7.1.3] 

4.7.1.3: Comparing two shot angles

All other possible combinations of camera angle variables were also tested but only two other significant differences emerged. The use of the rather dramatic shot angles of skew/canted and overhead revealed a highly significant (p<0.01) difference when comparing those advertisements aimed at boys and girls respectively. There were no significant differences, however, when these single-sex target audiences were compared with mixed audience advertisements. Finally, a significant difference (p<0.05) emerged when level and skew/canted angles were compared for girls’ and mixed advertisements, although there was no difference when comparing girls’ and boys’ or boys’ and mixed advertisements. These differences are summarised below [4.7.1.4]:

4.7.1.4: Comparing other shot angles

4.7.2 Testing product shots

After considering the use of camera angles, the next step was to test the length of shots used. The way in which a product or character is captured on screen is an important consideration. Indeed, in most cases at least, the target audiences are invited to engage with the images seen on screen, either in terms of relating to the character or making a positive connection with the product. For ease of reference, the lengths of the product shots counted are summarised below [4.7.2.1]:

4.7.2.1: Summary of product shot length

Both boys versus girls and girls versus mixed were tested but no significant difference emerged in either instance. The only difference was when boys’ and mixed audience counts were tested and this emerged as very highly significant (p<0.001). This calculation is summarised below [4.7.2.2]:

4.7.2.2: Comparing product shot length

4.7.3 Testing character shots

The various shot combinations were more complicated when considering the ways in which characters were framed on the screen. Indeed, the character could appear either singly or in mixed-sex combinations in all three shot lengths. Once again, for ease of reference, the content counts for character shots are summarised below [4.7.3.1]:

4.7.3.1: Summary of character shot length

Comparing character shots was an excellent illustration of why one should exercise caution when interpreting the results from statistical tests such as Chi-Square. Since there were many zero values, the variables were often untestable using the Chi-Square test. It was clear from simple observation, however, that there was an enormous difference between the way in which male and female characters appeared in the advertisements aimed at boys and girls respectively. Only token female characters appeared in the 43 boy-targeted advertisements while the same was true of male characters appearing in the 43 girl-targeted advertisements.

A number of statistically significant differences between shot sizes were also identified using the Chi-Square test. When comparing of the use of long shot, for example, an extremely significant difference (p<0.0001) emerged between girls’ and mixed advertisements. No significant difference emerged, however, when comparing boys’ and girls’ or boys’ and mixed advertisements. This calculation is summarised below [4.7.3.2]:

4.7.3.2: Comparing long character shots

Perhaps as one might expect, while level angles were the most frequently used, mid shots were the dominant character shots. These production conventions would appear to mark the ‘default form’ or ‘norm’ when constructing advertisements. When comparing the use of mid character shots here, extremely significant differences (p<0.0001) emerged in all instances. Perhaps the readings were made particularly dramatic by the zero values featured in the counts, but the results clearly emphasise the dramatic differences between the representation of characters in each audience category. These calculations are summarised below [4.7.3.3]:

4.7.3.3: Comparing mid character shots

Finally, the use of close-up character shots was compared. While there appeared to be no significant difference when comparing the use of close-up shots in boys’ and girls’ advertisements, one should stress that the zero values rendered these variables untestable. The actual contrast was clearly obvious. The use of close-ups in the comparisons of the other target audience contexts did, however, yield extremely significant differences (p<0.0001) in both cases. These calculations are summarised below [4.7.3.4]:

4.7.3.4: Comparing close-up character shots

4.7.4 Testing other production features

The possible combinations of the remaining production features were many and varied, but few yielded any statistically significant differences. Only one comparison emerged as significantly different (p<0.05) and that was the use of tilts in boy- and girl-targeted advertisements. This difference seemed particularly interesting in that the zero counts made the contrast (although only slight) a marked one. This calculation is summarised below [4.7.4.1]:

4.7.4.1: Comparing tilts

An alternative way to compare the data counts was to indicate whether certain features were present within a given advertisement or not. That is, to note whether or not the features ‘occurred’ in each advertisement, irrespective of frequency. When testing variables in this way, three comparisons emerged as significantly different. When comparing the presence or absence of the ped (either up or down) in girls’ and mixed advertisements, a highly significant difference (p<0.01) emerged. When the presence or absence of the skew or canted shot was compared for the same audience categories another very highly significant difference (p<0.001) emerged. Finally, when the presence or absence of the overhead shot in boys’ and girls’ advertisements respectively was considered, a significant difference (p<0.01) emerged. These calculations are summarised below [4.7.4.2]:

4.7.4.2: Comparing occurrences of camerawork features

4.8 Testing post-production and editing variables

4.8.1 Transitions

Perhaps one of the main ways to create a particular mood or level of pacing within a given televisual text is the use of transitions. In the context of the advertisements in this sample, one could argue that the use of transitions would work within the extremely condensed structure and format of advertisements to convey particular ideas about atmosphere and the nature of the product. For ease of reference, the transition counts are summarised below [4.8.1.1]:

4.8.1.1: Summary of Transitions

Through simple observation it becomes immediately obvious that there is a marked contrast between the way in which cuts and dissolves were employed in the boys’ and girls’ advertisements respectively. In terms of cuts, for example, the boy-targeted advertisements contained just over 300 more cuts than the same number of advertisements thought to be aimed at girls. The contrast in the use of dissolves was equally apparent, with the girls’ advertisements employing the technique more often than either boys’ or mixed audience advertisements. When these counts were tested, no significant difference was found between comparisons of boys versus mixed or girls versus mixed audience categories. However, there was an extremely significant difference (p<0.0001) when the boys’ and girls’ advertisements were compared. This calculation is summarised below [4.8.1.2]:

4.8.1.2: Comparing transitions

4.8.2 Shot duration

 Another interesting post-production or editing feature is that of shot duration which is directly related to the number of shots within each advertisement. The average duration of an advertisement tends to be about thirty seconds, although some advertisements may be anything up to sixty seconds in duration. In some instances, shorter advertisements have been increasingly used (Myers, 1999:124; Condry, 1989: 180 ff.). With this level of variation across advertisements in general, it was important to calculate average shot duration in the specific context of the toy advertisement sample. So the duration of each advertisement was timed in seconds and the number of different shots in each were counted. Details of duration and number of shots are included in the content summaries for each advertisement (Appendices C-E). The average length of shot within each target audience category was calculated, as summarised below:

·Average length of shot for boys’ advertisements: 1.23 seconds.

·Average length of shot for girls’ advertisements: 1.73 seconds.

·Average length of shot for mixed audience advertisements: 1.17 seconds.

From these target-audience figures an overall average length of shot was calculated as being 1.38 seconds. This figure was then used as a basis for comparing the advertisements in the sample in terms of whether the shot duration for each respective audience group was above or below the average. These counts are summarised below [4.8.2.1]:

4.8.2.1: Number of advertisement above or below average shot duration

It is clearly apparent that the majority of both boys’ and mixed audience advertisements in this sample were below average in terms of shot length, while the majority of girls’ advertisements were above average. One could therefore argue that the advertisements aimed at boys and mixed audiences were similar in terms of overall pacing. As anticipated, a very highly significant difference (p<0.001) emerged when comparing the average lengths of shot in advertisements for boys and girls. Comparing girls’ and mixed advertisements also resulted in a highly significant difference (p<0.001). Also, as expected, there was no significant difference between boys’ and mixed advertisements. These calculations are summarised below [4.8.2.2]:

4.8.2.2: Comparing shot duration

4.8.3 Voiceovers

Within the advertisement sample as a whole, many different voice types were heard. Indeed, one could argue that ‘voice’ includes all sounds articulated orally by human beings that, in this instance, included singing, chanting and any verbal exchanges between the characters appearing on the screen (men, women and children, either singly or in combination). In terms of the spoken word, there were also many variations in accent, tone and pitch. It was not an easy task to quantify or qualify such distinctions objectively and the richness in variation in the context of this advertisement sample would also translate into unnecessary complications and subtleties. Voiceover was therefore viewed in the strictest sense, referring to an invisible speaker who introduces or comments on the product being advertised. That is to say, the term voiceover was used to refer to the off-screen narrator for each product.

Classification of the advertisements fell into three categories of male, female and no voiceover. The male and female voiceover counts included text spoken by both adults and children. In those advertisements counted as having no voiceover, other ways were often utilised to convey the sales messages, such as the product jingle (sung text). Details of the voiceovers in each advertisement are included in the content summaries (Appendices C-E) and summarised below [4.8.3.1]:

4.8.3.1: Summary of voiceovers

Once again, strong patterns emerged from the counts before the Chi-Square test was applied. Perhaps most striking was the fact that no female voiceovers were used in the advertisements aimed at boys, while very few male voiceovers were used in advertisements aimed at girls. What also seemed significant was the fact that there was a predominant use of male voiceovers in the mixed audience advertisements. These trends were confirmed when tested since extremely significant differences (p<0.0001) emerged when boys’ versus girls’ and girls’ versus mixed advertisements were compared. Boys’ versus mixed advertisements, however, produced no significant difference when tested. The results are summarised in the following table [4.8.3.2]:

4.8.3.2: Comparing voiceovers

4.9 Testing other variables of setting and music

Setting and music, unlike the other variables discussed so far, are not listed in the content summary (Appendix B.3). One could argue that they are not immediately applicable to notions of manifest content, but I would suggest that both elements are significant and important in terms of contextualising the screen images and enhancing the overall ‘mood’ of the advertisements. For this reason, the variables will be discussed in detail here.

4.9.1 Setting

For reasons of workability, I tried to keep references to setting as clearly categorised as possible. The location categories of indoor, outdoor, nondescript, fantasy and specific were terms most often used. The term indoor referred to any scene set inside a house, garage or recognisable place with walls and a roof. In contrast, the term outdoor was used to refer to the park and garden settings as well as cityscapes, rugged areas of wasteland, countryside, beaches and any other open-air setting away from ‘home’. The term nondescript referred to settings that were not easily identifiable, particularly where the setting was omitted in favour of focusing on the product, or were the product was shot against a blank background such as a photo-shoot style backdrop.

The term fantasy described settings that appeared ‘other worldly’, such as fairyland or secret and hidden miniature worlds. This term also included the fictional worlds in which specific characters appeared, such as Batman in Gotham City. The final reference was specific settings. In this instance, the setting was vital to the process of contextualising the product, such as a football match, a racetrack, the Wild West, an operating theatre or on-board ship. The settings noted in the sample are summarised in Appendix B.4. Occasionally, more than one setting was used in an advertisement and these instances are included in brackets.

Perhaps one of the most noteworthy differences was the fact that more indoor settings were used in the advertisements aimed at girls. Interestingly, however, there was no real difference between the boys’ and girls’ advertisements in terms of the outdoor settings, where there was once thought to be a greater use of outdoor settings in advertisements directed at boys (cf. Smith, 1994: 329/333). The contrast in the use of nondescript settings was also interesting. While many of the boys’ settings could be classed as nondescript, settings in girls’ advertisements were too identifiable and specific to be placed in this category.

A number of statistically significant differences emerged when the data were tested. When comparing indoor and outdoor settings in boys’ and mixed advertisements, for example, the difference emerged as being close to highly significant (p<0.001). Comparing boys and a mixed audience in terms of indoor and fantasy settings produced a highly significant difference (p<0.01), although no significant difference emerged when comparing girls and a mixed audience in the same way. These results are summarised below [4.9.1.1]:

4.9.1.1: Comparing settings

4.9.2 Music

After listening to the audio soundtracks, a considerable list of music ‘types’ was generated which included rock, adventure-style, techno/dance, pop/disco, Country & Western, 1970s, classical, synthesised, slapstick, Caribbean, sea-shanty, jingle, film or television theme and sound effects. Each of these terms will now be defined in turn.

Rock described any music involving the sounds of an electric guitar. Adventure-style described music such as that heard in films like Indiana Jones or James Bond, to warn of imminent danger. Techno/dance described music with a heavy base beat, periodic synthesised sounds and rapid rhythm. Pop/disco was made distinct from techno/dance, in that the sounds were softer and more tuneful, yet still upbeat and ‘dancy’. This type of music may also be heard in the current Top 40 Pop Chart. Country & Western is self-explanatory, including high-speed violin and banjo playing. As a decade, 1970s was unmistakably characterised by tunes such as Saturday Night Fever that remain rather ‘groovy’!

Classical is self-explanatory, referring to orchestral-style music. Synthesised, on the other hand, referred to gentle synthesised background melodies that were neither tuneful nor instantly recognisable but nevertheless established a soothing mood for the whole advertisement. Slapstick referred to comical and quirky music, akin to that used in sketches by Monty Python. Caribbean-style referred to the sound of steel drums, while Sea-shanty referred to accordion-style music, akin to that associated with a pirate ship scenario. Product jingle referred to the songs written especially for particular products, incorporating explanatory lyrics or music consistently associated with certain products. Theme from film/T.V. series is also self-explanatory, in that the music was lifted directly from an accompanying television programme or film, where spin-off merchandising was being marketed. Sound effects referred to the additional noises that enhanced the product, such as screeching car tyres or crying babies. Not all advertisements in the sample contained an audio soundtrack, however, so these were noted as having no music. The music heard in this advertisement sample is summarised in Appendix B.5, which also notes any multiple music occurrences.

Many initially striking differences emerged. Rock music, for example, was frequently used in advertisements directed at boys, but appeared only twice in girls’ advertisements. A similar degree of contrast emerged in terms of product jingle, since many of the girls’ advertisements utilised this music type compared with only a small number of jingles in the boys’ advertisements. In addition, girls’ advertisements made greater use of synthesised music, while sound effects were frequently heard in boys’ advertisements. Very interestingly, no advertisements directed at girls used music from film or television programmes, suggesting that media parallels and cross-media merchandising may be more usually associated with products intended for boys. Overall, the greatest variety of music was heard in advertisements directed at a mixed audience.

When applying the Chi-Square test, a number of differences emerged. When comparing the use of rock and synthesised music in advertisements aimed at boys and girls, a very highly significant difference (p<0.001) emerged. Comparing boys’ and girls’ advertisements in terms of product jingle and theme music from film or television also revealed another very highly significant difference (p<0.001). These results are summarised below [4.9.2.1]:

4.9.2.1: Comparing music types

Some comparisons were listed as ‘invalid’ when tested, but if this warning is ignored a number of further differences between the music types can be noted. A significant difference (p<0.05) emerged when comparing adventure-style and synthesised music in boys’ and girls’ advertisements. A very highly significant difference emerged (p<0.001) when comparing the use of classical and slapstick music in advertisements aimed at girls and mixed audiences. Contrasts also emerged when comparing the use of classical and rock music in advertisements aimed at boys and girls respectively, where there was a very highly significant difference (p<0.001). These results are summarised below [4.9.2.2]:

4.9.2.2: Comparing music types (marked as ‘invalid’)

Marking the occurrences of different music types within the sample also worked particularly well in the context of music. When comparing the use of jingle versus no jingle, for example, differences were evident. When comparing boys’ with girls’ advertisements an extremely significant difference (p<0.0001) emerged. Similarly, when comparing girls’ with mixed advertisements, a very significant difference (p<0.001) emerged. These results are summarised below [4.9.2.3]:

4.9.2.3: Comparing music occurrences (i)

The degrees of difference in occurrences of rock versus no rock music were also tested. An extremely significant difference (p<0.0001) emerged when comparing boys’ and girls’ advertisements, while a very significant difference (p<0.001) emerged when comparing boys’ and mixed advertisements. These calculations are summarised below [4.9.2.4]:

4.9.2.4: Comparing music occurrences (ii)

A final comparison of occurrences was made which revealed marked differences in the use or otherwise of slapstick music in the advertisement sample. While no significant difference emerged when comparing boys’ and girls’ advertisements, a highly significant difference (p<0.01) emerged when comparing boys’ and mixed advertisements and a very highly significant difference (p<0.001) emerged when comparing mixed with girls’ advertisements. These calculations are summarised below [4.9.2.5]:

4.9.2.5: Comparing music occurrences (iii)

4.10 Discussion and possible implications

4.10.1 Production and camerawork features

A number of interesting patterns emerged from the counting and testing of production and camerawork features in this sample of toy advertisements. In terms of shot angles (high, low, and level), boys’ and girls’ advertisements were fairly distinctive. The boys’ advertisements tended to use both high and low angle shots more often than the girls’ advertisements did [4.7.1.1]. One might interpret this in different ways. It has often been suggested that high angles, for example, convey a superiority of status, where one looks down upon a scene (Goffman, 1979; Berger, 1991: 26). Extreme high level angles are often referred to as suggesting detachment (Millerson, 1985: 68/70). One could argue that the use of high angle shots reflected the so-called ‘masculine’ status of elevation and non-involvement.

Low angles, on the other hand, are conventionally interpreted as suggesting greater ‘potency’ in what is depicted. It forces one to look up to who ever or whatever is framed in this way (Zettl, 1999: 190; Messaris, 1997: 34-35; Kress & van Leeuwen, 1996: 146; Wurtzel & Rosenbaum, 1995: 44; Millerson, 1985: 68-69). Goffman (1979), in contrast, referred to low angle shots as creating a feeling of inferiority and reduced status (cf. Berger, 1991: 26). The boys’ advertisements also tended to use overhead shots more frequently, which often proved dramatic in their extremity. In terms of the more ‘unusual’ camera angle of the canted or skewed shots, the girls’ advertisements used the technique less often than the other target groups. This angle has been associated with dynamism, energy and activity (Zettl, 1999: 91; Zettl, 1995: 74) so girls advertisements, as a result of their lesser use of this technique, might be conveying none of the above.

When considering the occurrence of certain camerawork features, the girls’ advertisements differed from the mixed advertisements in that they employed peds (either up or down) more frequently [4.7.4.2]. The perceptual psychologist James Gibson suggested that ‘the moving camera… is the reason for the empathy that grips us in the cinema’ (Gibson, 1979: 298). Peds may therefore enhance the viewer’s sense of involvement. When comparing the use of tilts [4.7.4.1], a further contrast emerged between boys’ and girls’ advertisements that may once again reflect this notion of involvement and the ‘active’ following of on-screen events. One could argue that girls might be associated with a greater sense of (emotional) involvement.

No notable differences emerged in terms of the shot sizes used to depict product, since long, mid and close-up product shots were dispersed fairly evenly between the audience categories [Appendix B.3]. There was a difference in the counts between boys’ and mixed advertisements, but one could argue that this would be accounted for by the different number of ‘types’ (43 boys’ advertisements, 31 mixed advertisements) and the subsequent contrast in the number of shots. That is to say, there were far more shots in the sample of advertisements aimed at boys and therefore many more notable production features.

The boys’ advertisements emerged as significantly different when compared with the girls’ advertisements because they employed far more long shots featuring characters. Perhaps the main function of the long shot is to establish the scene by showing a complete view from some distance away. One could argue that the greater use of such shots might help to encourage field independence which, at its most basic level, refers to the ability to separate parts or details from a whole. Researchers have concluded that there is often a greater tendency in males than in females to be field independent (Witkin, 1970). Therefore, the advertisements in this sample might be offering the male viewers greater opportunity to deconstruct the composition of each shot by allowing them to ‘stand back’ and view the whole. The use of long shots may also suggest greater ‘distance’, in the same way that the high angle shots might suggest ‘detachment’; an integral non-involvement with the scene either physically or emotionally. These traits are arguably stereotypically ‘masculine’.

The boys’ advertisements were also significantly different from the mixed advertisements in that they made much less use of close-up character shots. Despite the fact that the difference between the boys’ and the girls’ advertisements did not reach the 5% level of significance, the girls’ advertisements were closer to the mixed advertisements in terms of character shot sizes. Close-ups are often said to mimic an ‘intimate’ face-to-face distance between oneself and the person depicted on the screen (Zettl, 1999: 190; Kress & van Leeuwen, 1996: 130-135). This type of shot tends to focus attention on a person’s feelings or reactions (Millerson, 1985: 61) or encourages empathy (cf. Singleton-Turner, children-media-uk archive). This interest in emotions is arguably a more ‘feminine’ tendency. Indeed, Modleski (in Baehr & Gray, 1996: 106) has argued that ‘close-ups provide the spectator with training in ‘reading’ other people’, which once again might be regarded as a stereotypically ‘feminine’ skill. According to Zettl (1999: 187) close-ups ‘intensify the event’ depicted on screen, creating a very different ‘feel’ when compared with the detachment and distance of long shots. Empirical studies have shown that close-ups lead to increases in both attention and involvement (Lombard, 1995; Reeves et al., 1992, both in Messaris, 1997: 29). 

Overall, however, the contrast in character shots between boys’ and girls’ advertisements might lie not so much in the size of shot but in the presence or absence of certain characters according to the target audience. The boys’ advertisements in the sample featured only two on-screen female characters (one girl, one woman), while the girls’ advertisements featured males in eight shot instances [4.7.3.2-4]. This may have little to do with notions of detachment or emotional involvement but might be rather more to do with simple psychology. Advertisers have acknowledged that girls are more comfortable in ‘accepting’ male on-screen characters than vice versa (cf. Acuff, 1997: 157; Clark, 1988: 190). While one of the main purposes of advertising is to ‘sell’, it might be regarded as ‘safer’ to use the more masculine form to gain ‘acceptability’ across all audience categories.

4.10.2 Post-production and editing features

In relation to the use of transitions (cut, dissolve, fade and ‘swirl’) in this advertisement sample, the emergent patterns did seem to support existing research in the field. Cuts were by far the most common transition used across all audience categories but those advertisements aimed at girls used more dissolves than those aimed at boys. Millerson (1985: 111) suggested that the strength of the cut technique is the powerful impact that it can have on the audience due to the ‘sudden change’ signified. Perhaps the harsh abruptness of cuts might be more ‘masculine’ in feel, creating a sense of assertion and action (Wurtzel & Rosenbaum, 1995: 445; Millerson, 1985: 115; Huston et al., 1984: 708; Welch et al., 1979: 206). While the conventions of ‘invisible (continuity) editing’ may seek to make cuts unobtrusive, one could argue that they still have the power to subconsciously affect the way in which a viewer reacts to a sequence of events on screen. In stark contrast, it has been suggested that the ‘softness, gentleness, predictability and slow gradual change’ of dissolves connotes passivity (Welch et al., 1979: 207), which might in turn create a more ‘feminine’ feel to a sequence of shots. Even young viewers are said to recognise the gender connotations of such coding (Huston et al., 1984: 714). It would be extremely interesting to conduct further study on the way in which viewers perceive the use of certain transitions over others and the possible audience impact of one convention over another. As yet, few researchers have investigated this issue in detail.

In terms of shot duration, the findings here seemed to correspond with existing research. More shots tended to be used in advertisements aimed at boys and consequently the average duration of each shot was shorter. However, such a difference would be unlikely to be obvious to those who view the advertisement texts because they occur in only fractions of a second. It is still possible to argue that the very existence of such subtleties within the framework of an advertisement text is significant in terms of viewer response – an issue that will be investigated later during interview sessions with young children.

In the context of the cutting rates used in film and television productions, a pattern emerged when comparing the boys’ and girls’ advertisements in this sample [4.8.2.1]. The cutting rates seen in the boys’ advertisements were very rapid compared to slower pacing in the girls’ advertisements. In this context, one might suggest that fast cutting rates were stereotypically ‘masculine’ and that the cutting rates in general could be described as ‘masculinised’. Rapid pacing is often noted as one of the perceptually salient features of advertisements which serve to attract and maintain children’s attention (Welch et al., 1984; Wright et al., 1984; Meyer, 1983). Indeed, this notion formed the working theory behind such children’s programmes as Sesame Street, in which information was presented in brief ‘sound bites’. Experimental research by Penn (1971) has also shown that rapid cutting led viewers to rate films highly in terms of ‘potency’ and ‘activity’ (cf. Heft & Blondal, 1987; Hochberg & Brooks, 1978: 291-295). Furthermore, film-editing theorists refer to the ‘expressive’ function of rapid cutting in terms of building dynamic moods such as ‘excitement’, ‘intensity’ and ‘tension’. Rapid cutting is consequently juxtaposed with slower cutting rates which are said to create feelings of ‘tranquillity’, ‘calm’ and ‘relaxation’ (Zettl, 1999: 250/256; Brandt, 1994; Zettl, 1992: 349; Crisp, 1987). One might therefore posit that the rapid pacing of the advertisements aimed at boys in this sample reflected the stereotyped notion of masculinity as action-oriented.

Finally, the use of voiceovers here also seemed to correspond with other research findings. More male than female voiceovers were used overall, even in those advertisements aimed at girls [4.8.3.1]. On no occasion were female voiceovers used in advertisements aimed at boys, while there was a predominance of male voiceovers in mixed advertisements. It has often been argued that the high proportion of male voiceovers used in advertisements adds a sense of authority to both the product and the gender of the voiceover (Fowles, 1996: 208-9, 211). This pattern would seem to follow the theory posited by Welch et al. (1979: 207), that males are given authority in all aspects of content other than that which is distinctly female. Interestingly, the sample revealed that there were many more female than male voiceovers in the advertisements directed at girls. This is contrary to the use of voiceovers in adult-directed advertisements (cf. Courtney & Whipple, 1983: 17) and would appear to mark a wholly different approach to children. A departmental colleague identified the intriguing contradiction by asking whether there was a point at which ‘girls’ suddenly become ‘women’ in the world of advertising, and it would certainly be interesting to pursue this further.

4.10.3 Other production features

As well as the formal camerawork and editing features, setting and music were also considered. Other studies in the field of gender and advertising have revealed differences in the use of location in accordance with gender. Craig (1992: 206), for example, noted that female characters were more likely to be advertising products set ‘indoors at home’, with men more likely to appear in settings away from the home or in business locations. This observation once again raises issues of ‘importance’ and ‘authority’, where men would appear to have the upper hand. In the context of this study, the settings used in the advertisements aimed at girls were more likely to be indoors in the home environment than those advertisements aimed at boys [Appendix B.4]. One would assume from this that the boys’ advertisements would be more likely to be set outdoors, but when the outdoor settings for boys’ and girls’ advertisements were compared, there was little difference.

The main degree of setting contrast between the male and female advertisements seemed to be between the use of ‘nondescript’ and ‘identifiable’ settings (cf. Smith, 1994: 333). A significant number of the boys’ advertisements could be coded as nondescript and one could argue that this would equate with freedom of imagination. In other words, the male audience seemed to be placed in a position whereby they could choose their own location for toy-play and not be restricted by specific screen images. No advertisement aimed at the female sector of the audience had a setting that could be coded as nondescript because all the locations were identifiable and ‘everyday’, such as bedrooms, sitting rooms, bathrooms and kitchens. One could argue that such specific settings could be restrictive in the sense that they are ‘grounded’ and do not provide the option of fantasy. A significant difference was also noted in the use of indoor and outdoor settings in boys’ and mixed advertisements [4.9.1.2]. One could argue, however, that this had more to do with the nature of the products being advertised in the mixed advertisements, such as family board games designed to be played with indoors, rather than the target audience itself.

Music could also be described as a significant contributor to the overall feel of an advertisement in that it is able to conjure specific moods and attitudes. Courtney and Whipple (1983: 22) made brief reference to the music used in children’s advertisements by explaining that boys’ advertisements often employed loud music while girls’ advertisements were more likely to use gentle background music. A number of interesting patterns seemed to emerge in the context of this investigation. Through simple observation of the music counts [Appendix B.5] there was a clear predominance of rock music in the boys’ advertisements compared to the principal use of the product jingle in the girls’ advertisements. In addition, the girls’ advertisements tended to use synthesised music while the boys’ advertisements employed more sound effects. One could argue that these music types reflect the stereotyped notions on ‘gentle’, ‘soothing’ and ‘calming’ as being suited to the more ‘feminine’ while ‘loud’, ‘action-packed’ and ‘in-your-face’ were features considered more ‘masculine’ 

Another interesting pattern regarding music types emerged during a comparison of occurrences [4.9.2.3-5]. The girls’ advertisements mostly used product jingles while the boys’ advertisements used rock music, and mixed advertisements used slapstick music. While rock music and product jingles were already identified as having connotations of masculinity and femininity respectively, the use of slapstick music raised new questions in the context of mixed advertisements. The frequent use of slapstick music in this context may be because it arguably appeals to all children’s sense of the absurd and comical, irrespective of sex (cf. Acuff, 1997; Del Vecchio, 1997). It should also be noted, however, that slapstick music was not a prominent feature in those advertisements aimed primarily at boys or in those aimed at girls, prompting the question of whether the advertising of gender-specific products is a more serious business than the promotion of toys with universal appeal.

4.11 ‘Marked’ and ‘unmarked’ categories

At this point, one feels that it is appropriate to introduce a concept derived from linguistics and frequently used in semiotics concerning the theory of ‘marked’ and ‘unmarked’ categories (see Chandler, 1994a). Constructing an advertisement (or any other text) involves a multitude of stylistic choices between alternative ‘paradigms’ within the various technical ‘codes’ of the medium and the genre. Televisual editing code paradigms might include the use of a certain transition (such as cut or dissolve), whilst camerawork code paradigms might include the use of a certain shot length (long, mid or close-up). Where there are clear-cut alternative paradigms, where one paradigm can be clearly substituted for another, the form most commonly used can be referred to ‘unmarked’. Alternative forms are referred to as ‘marked’ because they are made conspicuous by their ‘unusualness’. Furthermore, these paradigms are awarded different values in that the ‘unmarked’ often seems ‘normal’ and ‘natural’ while the ‘marked’ is ‘different’ and ‘unnatural’. In terms of the music used in this advertisement sample, for example, the use of rock music would be ‘marked’ in girls’ advertisements while the use of product jingles would be ‘marked’ in boys’ advertisements.

Similar trends of ‘marked’ and ‘unmarked’ categories were observed in other content analysis categories where differences were noted in terms of low angle shots, level shots, canted shots, peds up and down, cuts and dissolves, shot duration, and the use of male voiceovers. It should, however, be stressed that this content analysis dealt not only with advertisements aimed at boys’ and girls’ but also those aimed at a mixed audience. As a result of the so-called dual appeal of the mixed audience advertisements one could arguably expect that they would mark the middle ground or ‘norm’ of product representation. It would therefore be reasonable to assume that the use of formal features in this mixed audience category might tend to reflect an intermediate position between the counts of each feature in boys’ and in girls’ advertisements respectively. Furthermore, an advertisement could be described as ‘marked’ if it diverged from the ‘norm’ of the mixed audience advertisement.

The content analysis findings, however, did not consistently reflect this pattern of ‘mixed’ as ‘norm’. The table below [4.11.1] provides a convenient summary of the way in which the mixed audience advertisements could be applied as the ‘norm’ for some features but not others. The left-hand column lists those technical featured that emerged as ‘unmarked’ because the ways in which they were used in the sampled boys’ and girls’ advertisements were consistent with the mixed advertisement patterns. The right-hand column, in contrast, lists all those features that were used differentially when the boys’ and girls’ advertisements were compared with the mixed audience category.

4.11.1: Applying boys’ and girls’ advertisement features to the ‘mixed norm’

As a general rule the mixed audience advertisements in this sample were more closely aligned to the boys’ rather than the girls’ advertisements. One could therefore argue that the similarity of the mixed advertisements to those for boys ‘normalised’ the ‘male’ by presenting the female as ‘other’. That is to say that fast-paced, loud advertisements with extreme camera angles were ‘normal’ while slower-paced, soothing and gentle advertisements were ‘female’ (cf. Walkerdine, in Curran et al. 1996: 325). A useful summary of those features for which the ‘female’ category emerged as ‘marked’ is given below [4.11.2]. When compared with the other audience categories these differences in approach are made all the more dramatic.
 

4.11.2: Summary of content features marking gendered advertisements as different

From the above table, one could argue that a number of content features emerged as distinctly ‘female’. Girls’ advertisements were marked in terms of their lesser use of extreme shot angles, long product shots, male character shots and rock music. They also emerged as marked for their greater use of dissolves, female voiceovers, product jingles and longer shot duration. The way in which the girls were highlighted as ‘other’ is still more dramatic because the boys’ and mixed advertisements were so closely aligned. Again, the debate may have turned full circle when one considers that advertisers, by their own admission, will consistently pitch their commercials at the ‘masculine’ because this approach is more likely to be universally accepted and therefore more likely to be successful (cf. Acuff, 1997: 157; Clark, 1988: 190).

Finally, one should return briefly to the hypotheses formulated at the beginning of this chapter in order to determine whether the results of this analysis support the existing content analyses findings in children’s advertising. In terms of the variety of observed camera patterns (Hypothesis 1), it would appear that the girls’ and boys’ advertisements in this sample contained an equally varied style of presentation, with only five zero counts in advertisements for both audience categories respectively [Appendix B.3]. It is therefore inappropriate to suggest that the advertisements aimed at boys in this study contained greater stylistic variety than those aimed at girls. My findings in respect of the use of transitions (Hypothesis 2) clearly support the existing research in this field, in that more fades and dissolves were used in the advertisements aimed at girls [4.8.1.1]. This analysis demonstrated that the girls’ advertisements stereotypically reflected the consensus that females are assumed by advertisers to prefer gradual and gentle transitions, while males are assumed to favour more aggressive and abrupt changes of scene.

There were more male voiceovers in advertisements aimed at boys and mixed audiences in this sample (Hypothesis 3). However, in stark juxtaposition to the use of male voiceovers in adult advertisements directed at women (cf. Pesch et al., 1980; Hennessee and Nicholson, 1972; Barcus, 1971; Dominick and Rauch, 1971, all in Courtney & Whipple, 1983), there were many more female than male voiceovers in advertisements directed at girls. In the case of the children’s toy advertisements in this sample, male voiceovers perhaps only dominated because there were no female voiceovers whatsoever in the advertisements aimed at boys [4.8.3.1]. In terms of other auditory features (Hypothesis 7), the findings from this study showed that the boys’ advertisements made far greater use of sound effects than those advertisements aimed at girls. The music in the boys’ advertisements was also much louder, with more instances of rock music than any other group [Appendix B.5]. Therefore, my findings once more seemed to support previous research in this field.

It is possible to consider the issues of shot number (Hypothesis 4) and shot duration (Hypothesis 5) together because one can assume that the average duration of a shot will also reflect the number of shots seen in each advertisement. This would be due to the fact that there is little variation in the overall length of each advertisement in this sample. Clearly, therefore, my findings support the existing research, in that there were more shots used in advertisements directed at boys where the average duration of each shot was shorter than in those advertisements for girls [4.8.2.1].

In addition to shot number and duration, the number of advertisements for each audience type was also considered (Hypothesis 8). The number of different kinds of advertisements (‘types’ rather than ‘tokens’) did not differ in this sample, since there are forty-three types for both boys and girls. However, the frequency with which these were shown was where the real contrast became startlingly obvious. The boys’ advertisements appeared 132 times, while the girls’ advertisements appeared only 94 times [4.2.1]. This difference, and the fact that boys’ and mixed advertisements are often stylistically similar, seemed to give the impression that there were more advertisements aimed at boys than at girls.

Finally, the settings used in each advertisement were considered (Hypothesis 6). It was undeniable that, through being set indoors, the majority of the girls’ advertisements in this study did seem to be based in the home [Appendix B.4]. While indoor settings for boys often included more exciting scenes like warehouses and sheds, indoor settings for girls were generally restricted to the rooms in a house. In this sample, girls seemed to be restricted to a domestic arrangement, echoing the use of the settings in advertisements aimed at an adult audience (see Chapter Two). However, one should not neglect the fact that there was little difference between the boys and girls in terms of outdoor settings. It is therefore an exaggeration to claim that girls are always seen in the home. There was perhaps a greater contrast between the boys’ and the girls’ advertisements in terms of the abstract notions of ‘fantasy’ and ‘reality’. It could be argued that the girls’ advertisements were rather restricted to ‘reality’ in the sense that the settings were always identifiable, while the boys’ advertisements may have afforded more ‘fantasy’ play by often being grounded in unidentifiable settings. This finding may imply that while girls are restricted, boys are given greater freedom (and, arguably, power).

Overall, undertaking a content analysis of the toy advertisements proved to be a very interesting and valuable exercise. Not only did the structures of these advertisements reveal interesting patterns of audience address, but they also went some way to answering the research issues. Considering that the last content analysis to be undertaken on children’s advertisements (in Britain) was by Manstead & McCulloch (1981), a subsequent decade of approach and address has seen little change. As the ‘marked’ category, girls’ advertisements are still built upon recognised ‘female’ structures. Indeed, the responses from the adults who coded the target audiences also substantiated this notion.

Perhaps most interesting was the fact that it was often difficult to discriminate between the approaches used in many of the advertisements directed at a mixed audience, since the nature of address was rather ‘masculine’ while the product itself would certainly be suitable for both sexes. In this sense, there is a strong connection between the ways in which products are presented in advertisements aimed at adults and children. While adult advertisements showed a greater proportion of male over female voiceovers, for example, children’s advertisements seemed subtler. While they made good use of female voiceovers for products regarded as ‘female’, there was an underlying sense that products given a male voiceover were some how ‘better’ or more desirable.

The content analysis focused of both the production and post-production features employed in the toy advertisement sample, even though only studies of the post-production features tend to appear in the published research (e.g. Huston et al., 1979). Overall, my analysis of the post-production features supported the existing research in the field, in terms of revealing certain definable gender differences. More importantly, however, my analysis of the production features both echoed and reinforced existing (gendered) patterns, making the theory of specific ‘technical gendering’ in advertisements more concrete and convincing. Gender stereotyping in advertisements is such a consistently well used technique that it seems difficult to imagine any other form of instantly recognisable address. It therefore seems unlikely that advertising agencies will ever ‘move with the times’ to reflect social change (cf. Doolittle & Pepper, 1975: 140). It is impossible to answer with certainty the question of whether or not the construction of advertisements mould and sustain the notions of ‘gendered’ consumer stereotypes because the roots of the issue lie far deeper within the heart of our society. It would be rather more illuminating to consider the ways in which young children ‘negotiate’ meanings from such stereotyped images (see Chapter Seven).

Summary

The process of target audience classification yielded a strong intercoder reliability level of 98.3%. The boy- and girl-targeted advertisements were perceived by the coders to form two distinctive categories and were not seen to overlap on any occasion. The ‘mixed’ classification was used when the target audience seemed less clear-cut. When the ‘tokens’ and ‘types’ counts were considered, it was notable that those advertisements deemed to be aimed at boys were shown more frequently than those for girls, perhaps fostering a sense of ‘male domination’ during a typical commercial break.

Through the counting and coding of specific formal features, a number of initial observations could be made. For example, the level shot was the most frequently used shot angle, while the boys’ advertisements employed more low-angle shots. Male characters appeared more frequently in the girl-targeted advertisements than vice-versa, while the mixed advertisements generally featured more male than female on-screen characters. Furthermore, the girls’ advertisements made greater use of dissolves while the boys’ advertisements more often used cuts.

Strong patterns emerged when the production feature counts were formally tested using the Chi-square test. The boys’ advertisements made greater use of extreme camera angles (high, low and overhead shots), while the girls’ advertisements made greater use of peds. There was no significant difference in the shot sizes used to show the products being advertised but shot sizes featuring characters differed in that the boys’ advertisements used more long shots compared to the greater use of close-ups in girls’ advertisements. The girls’ advertisements used the most number of dissolves, while the boys’ advertisements featured more cuts. Shot duration in both the boys’ and mixed audience advertisements was below the average shot duration for the sample as a whole, while the shot duration in the girls’ advertisements was mostly above the average. The cutting rates were accordingly either faster- or slower-paced for each target audience.

Perhaps the most striking difference in post-production features was the predominance of male voiceovers across the sample as a whole. Rock music and sound effects were employed most often in the boys’ advertisements, while the girls’ advertisements tended to feature product jingles and synthesised music most frequently. Slapstick music was featured almost exclusively in the mixed audience advertisements. There was little difference, however, between the types of settings used in the advertisements. This was the only clear contradiction between the patterns in this advertisement sample and those identified by other researchers in the field (cf. Smith, 1994: 333; Peirce, 1989, and Dominick & Rauch, 1971, both in Courtney & Whipple, 1983: 17). From comparative testing of the content features and using the mixed categorisation as the ‘default form’ or ‘norm’, the girls’ advertisements can be described as the ‘marked’ category in terms of both construction (form) and content. Each of the content features identified and tested in this analysis might arguably carry subtle gender connotations that could be subconsciously internalised by the (child) audience.

▲

This page was last modified 18 Apr 2006