1
Taking Advertising Literacy to a Higher Level:
An Exploratory Multilevel Analysis of
Children’s Advertising Literacy
Pieter De Pauw, Ghent University, Belgium
Verolien Cauberghe, Ghent University, Belgium
Liselot Hudders, Ghent University, Belgium
Corresponding author:
Pieter De Pauw
Dep. of Communication Studies (Ghent University)
Korte Meer 7-9-11, 9000 Ghent, Belgium
(E)
(T) +32 9 264 97 58
Pieter De Pauw is a PhD candidate at Ghent University and would like to be nominated for the Best Student Paper Award.
1
Taking Advertising Literacy to a Higher Level:
An Exploratory Multilevel Analysis of
Children’s Advertising Literacy
abstract
As few studies focus on how children’s coping with advertisingis affected by their environment, the present study uses multilevel analysis to explore the role of both primary (i.e. parents) and secondary socializing agents (i.e. classmates, teachers) in children’s advertising literacy. The results show that children’s cognitive advertising literacy and attitudes toward advertising are to a large extent determined by class-level processes. Their moral advertising literacy is a more individual matter, yet greatly influenced by children’s teachers – though in a remarkable manner. Parents do not seem to have much of an impact, except through socioeconomic factors such as educational background.
Purpose of the study
As parents and teachers must know, advertising has an apparent social dimension: on a regular basis, they experience children talking (among each other) about advertising and the brands and products they promote. In this light, it is somewhat remarkable that children’s coping with advertising has traditionally been investigated by focusing primarily on the individual child, and considerably less on their environment. Exceptions are the many studies on parental advertising mediation (e.g. Buijzen & Valkenburg, 2005), though these are scarce as concerns the contemporary advertising formats (e.g. Vanwesenbeeck et al., 2016). Recently, there has also been a surge in research on advertising literacy education at school (e.g. Nelson, 2015), though this is not about the daily impact of teachers through (whether or not deliberately) transferring advertising-related knowledge and attitudes on their pupils. Least studied in this regard is the influence of classmates, despite the findings that children’s processing of advertising is highly determined by their susceptibility to peer influences in the case of both TV advertising (ValkenburgBuijzen, 2005) and online social games(Rozendaal et al., 2013).
In other words, there is a clear need for more researchaboutcontextualeffects on children’s coping with contemporary advertising. Therefore,the present study exploreshow and to what extent children’s advertising literacy (and its cognitive, affective and moral components) is influenced not only by their individual and parental background features, but also by their classmates’ and teacher’s characteristics (related to advertising literacy). New compared to existing research, is that multilevel analysis is used to allow for simultaneously studying individual (children, parents) and class level (classmates, teachers) effects, and their interactions. In this way, the study identifies several macro processes (e.g. pro-advertising culture) that affect children’s advertising literacy over and above the effects of analogous individual level variables (e.g. own judgment of advertising).
Children’s Advertising Literacy
According to Hudders et al. (2017), advertising literacy can be defined as consumers’ advertising-related knowledge and skills, and the ability to recognize advertising and to critically reflect on it. The authors perceive advertising literacy as a multifaceted concept, entailing a purely cognitive, a largely affective, and a moral dimension.
The cognitive dimension of advertising literacy basically consists of the ability to recognize advertising (i.e. to distinguish it from non-commercial content) and to understandits commercial intent(Rozendaal et al., 2011). The present study focuses on the former, as recognizing advertising is the ultimate precondition for consciously processing ads.
The affective component of advertising literacy, then, refers to the conscious awareness of one’s initial, spontaneous emotional reactions toward advertising and abilities to regulate these emotions.However, as there is no measure yet following this definition, the present study concentrates on children’s attitudes toward advertising in general, which matches the recently emerged concept of ‘attitudinal advertising literacy’(Rozendaal et al., 2016). The rationale for this concept is that children who have a negative view on advertising will be more likely to resist ads when recognized as such, as a kind of ‘affective defense’. It should be noted that these attitudes are recently found to have an important moderating role, channeling the effects of children’s cognitive advertising literacy on ad effects (De Pauw et al., forthcoming). Therefore, the present does not only consider the attitudes toward advertising as a second outcome, but also has special attention for its possible interactions with the cognitive and moral processing of advertising.
Finally, the moral dimension of advertising literacy can be defined as the ability or readiness to reflect morally on advertising and its tactics. As argued by Friestad and Wright (1994), it is necessary to evaluate an ad and its tactics in terms of appropriateness (e.g. by reflecting on the fairness of using implicit marketing techniques), if the goal is to arrive at a well-substantiated, critical evaluation of the ad (and appropriate coping behavior).
Being exploratory, the present study attempts to explain as much variance as possible in these outcomes by linking them to variables connected to (advertising) literacy in the extant literature (see methods for an overview) – both on the individual level (child, parents) and thegroup or class (classmates, teacher) level. In this way, the analyses arrive at the best fitting model for each dimension of advertising literacy (cf. Gonzalez-Valenzuela et al., 2016). Accordingly, these dimensions are also considered as each other’s predictors, to gain insight in how they are interrelated.
Methods
A survey was conducted in 22 classes from 12 primary schools among 4th and 6th grade pupils (N = 392, Mage = 10.26, 43% girls), their parents (N = 188) and their teachers (N = 22). An extra dataset was created by aggregating the children’s data per class, and the resulting four datasets were linked to allow for multilevel analysis.
Children’s Questionnaire
Concerning the outcome variables, cognitive advertising literacy was measured by asking the children ‘How easy do you find it to recognize [format x]?’ (1 = very hard; 5 = very easy). This question was repeated for 5 contemporary ad formats (TV commercials, product placement, advergames, online banners and pre-roll video ads; each preceded by a brief explanation), to arrive atan index (M = 3.77 on 5, SD = 0.65). Attitudes toward advertising was operationalized by asking them ‘How much do you like advertising?’ and ‘How much do you find advertising annoying?’ (1 = not at all; 5 = very much) (M = 2.87 on 5, SD = 1.09). Moral advertising literacy involves three questions: ‘How often do you think about whether advertising…’ ‘…is honest?’, ‘…is misleading?’, and ‘…shows things like they really are?’ (1 = never; 5 = very often) (M = 2.85 on 5, SD = 0.96).
As regards the independent variables, i.e. the factors related to advertising literacy in the extant literature, three socio-demographic variables are included: children’s gender, age and media use (hours/day in free time watching TV and surfing the internet).As a cognitive predictor, awareness of the ad formats’ existence is measured by asking the children ‘Did you already know that [format x] existed (before we told you this)?’ (1 = not at all; 5 = certainly) (M = 3.90 on 5, SD = 0.70). The attitudinal factors adopted in this study are liking the advertising formats: ‘What do you think of [format x]?’ (5-point semantic differentials: ‘I find [format x]…’ ‘not nice – nice’, ‘not interesting – interesting’, and ‘not annoying – annoying’) (M = 2.75 on 5, SD= 0.68); and evaluating the ad formats in terms of appropriateness (I find [format x]…’ ‘dishonest – honest’ and ‘wrong – good’) (M = 3.23 on 5, SD = 0.72). Finally, children’s strategies to cope with advertising are assessed through four separate items, asking them: ‘When you see advertising, how often…’ 1) ‘…do you think about that advertising?’ (reflecting on ads)(M = 2.64, SD = 1.12); 2) ‘…do you try to avoid that advertising? (avoiding ads)(M = 3.39, SD = 1.27)’; 3) ‘…do you want to have the advertised product?’ (advertised product desire)(M = 2.49, SD = 1.00); 4) ‘…do you get a good feeling about the brand or product?’ (good feeling about the advertised brand or product) (1 = never; 5 = very often) (M = 2.44, SD = 1.05) (M = 2.44, SD = 1.05).
To measure peer influence, these individual-level variables were aggregated to summarize the mean score per class. Also adopted is children’s school grade (0 = 4th grade, 1 = 6th grade) and class SES as indicated by the percentage of children having parents with a master degree or higher (see parents’ questionnaire below) (0 = below median, 1 = above median).
Parents’ and Teacher’s Questionnaires
These questionnaires are largely similar to the children’s survey, except for the addition of following measures: parents’ educational attainment (1 = master/PhD vs. other = 0) and family size (number of children) as socio-demographics; and parents’ (and teacher’s) discussing of the new advertising formats with their children (pupils) to assess advertising mediation: ‘How often do you talk with your child (pupils) about the fact that…’ ‘…the news messages or blogs they read are sometimes advertising?’; ‘…sometimes brands or products are hidden in the TV programs, series and movies they watch?’; ‘…sometimes advertising is present on the websites they surf?’; ‘…the online games they play are sometimes advertising?’; ‘…the advertising they get to see prior to movies on the internet are sometimes based on their personal online behavior?’ (1= never; 5 = very often) (Mparents = 2.42, SD = 0.83; Mteachers = 2.80, SD = 0.57).
Statistical analysis
First, correlations are explored between the outcomes and the other study variables. Second, the significantly correlating variables are analyzed via multilevel techniques, to check whether the relationships still hold when accounting for the nested data structure, and to examine the separate contribution per predictor (block). Additionally, possible interaction effects are explored. Third, the study works toward more integrated multilevel models, in order to explain the observed variance per outcome, and to determine the combined contributions of the predictors (significant the in previous step). As these elaborate analyses (repeated for the three outcomes) yieldtoo much information to cover in this 5-page summary, the following section will only discuss the most prominent results.
Results and Discussion
Preliminary analyses show a relatively high proportion of the variation in cognitive advertising literacy (12%) and attitudes toward advertising (13%) occurring at the class level. In other words, there are great similarities in these outcomes between children belonging to the same class. While this is not the case for moral advertising literacy (1%), remarkable teacher influences on the individual child are revealed for this outcome(see below).
As regards cognitive advertising literacy specifically (Table 1), it is found that children are more skilled at recognizing the contemporary advertising formats when they are betteraware of their existence, and as they are more vigilant in avoidingthe ads they encounter. Additionally, it is shown that these awareness and avoidance effects are moderated by the extent to which children like the formats. In particular, it is found that the effect of awareness on recognition increases as children like the formats better. Furthermore, it is revealed that liking the formats can impede recognizing them, though only among children with a general tendency to avoid advertising. These interaction effects suggest that a certain ‘openness’ toward advertising may smoothen the path for cognitive advertising literacy to be developed and refined. The class context is also found to be an important mediating and moderating factor. In particular, it is demonstrated that the ad format awareness in class strengthens children’s own awareness, while the effect of the latter on recognition declines (or is supplemented) by the awareness present in the class. Interestingly, parents and teachersseem to have little influence here.
Concerning the attitudes toward advertising (Table 2), the analyses show that children are more positive toward advertising in general when they like the studied formats better – which is no surprise when considering the increasing emphasis on entertainment and emotional appeals in contemporary advertising directed at children (Wicks et al., 2009). On the group level, their attitudes seem to be influenced most strongly by their peersor classmates. Moreover, most of these class effects are contextual, i.e. having an influence over and above the individual level analogues of these variables, making it safe to speak of a real class ‘culture’ regarding advertising-related attitudes. In particular, it is found that children like advertising more when their classmates use less media, like the ad formats better and are more likely to have positive emotions when confronted with advertised brands or products - in other words: when their class is characterized by a weaker consumer socialization (due to less experience with advertising) (see John, 1999) and a pro-advertising culture. A positive attitude toward advertising is also more prevalent among children with a higher number of siblings, and most significantly, among children from classes characterized by a ‘lower’ parental educational attainment. Former result may be explained by a greater family size leading to lower parental involvement per child, resulting in lower literacy levels (Downey, 1995). As regards the latter, the less highly educated parents are perhaps less skeptical toward advertising, which is transferred onto their children, contributing to a pro-ad class culture (and consequently influencing each individual child). In general, children’s attitudes toward advertising seems hardly affected by teacher advertising literacy (and related factors) and mediational efforts.
Conversely, regarding moral advertising literacy (Table 3), it is the teacher that exerts the most determinative contextual influences, and the direction of these effects is quite remarkable. More specifically, the analyses show that children reflect more often on advertising’s appropriateness when their teacher is less skilled in recognizing the contemporary formats, and less often talks with them about the newer advertising formats. Moreover, it is only in latter condition that children’s awareness of the contemporary formats fuels their moral advertising literacy. Vice versa, these findings might indicate that teachers who are more advertising literate and more involved in children’s coping with advertising, tend to portray advertising in negative light in class. This might impede children’s openness to advertising, and ultimately desensitize their moral radar (in favor of reactance against advertising).
CONCLUSION
The main goal of this article is to explore both individual and group predictors of young children’s advertising literacy (and its cognitive, affective and moral dimension) for the contemporary advertising formats, using multilevel techniques to analyze data collected among the children themselves, their parents, classmates and teachers. As these analyses yield much information, this 5-page summary concludes with the most prominent considerations.
Perhaps the most important conclusion is that children’s coping with advertising has a significant social dimension, which cannot be ignored. As children’s advertising literacy is greatly influenced by their daily social context, it is strongly recommended to take group influences into account. Especially when conducting quantitative analyses on data collected in classes, it is necessary to statistically acknowledge that children are no independent subjects, but meaningfully nested within classes (which have significant effects on their respective members, and vice versa). This does not only guarantee more correct estimates and robust findings, but also allows to explain a great deal of children’s differences in advertising literacy that cannot be covered when only considering individual level variables. However, even if the researcher is not interested in the contextual effects per se, analyses should at least control for the grouped structure of the data (e.g. by adding a dummy for each class).
Multilevel analysis is indispensable, though, when the researcher does focus on contextual influences. The current study shows that such an endeavor is well worth it, as it appears to have revealed a veritable culture regarding advertising, and advertising literacy in particular, which cannot be overlooked. More specifically, following the definition of culture, classes seem characterized by a common set of shared set of meanings, understandings, cognitions, beliefs, etc. (see Van Houtte, 2005) which influence the individual members of these social units. In fact, the different dimensions of advertising literacy appear to be affected by different cultures, as constituted by children’s own background, peers and/or teachers. Moreover, the analyses uncover more nuanced cross-level interactions, in which the effects of individual predictors are modified by the characteristics of different social actors.
First, it is found that although children’s cognitive advertising literacy is primarily determined by their own awareness of the existence of the contemporary advertising formats, this individual effect can be strengthened and even supplemented by the awareness in class – attributing their peer group an empowering quality. As the current study seems to correspond to educational research showing that children’s formally induced literacy knowledge is practiced and fine-tuned trough peer interaction (Corsaro and Nelson, 2003), future research testing educational interventions could take advantage of such macro processes to more effectively impart knowledge and stimulate awareness of ad formats and tactics.
Second, as regards children’s attitudes toward advertising, the presence of a class culture is most pronounced. In particular, many contextual effects are found to emanate from the peer group, showing that children are more positive toward advertising when their classmates have less media experience and are more compliant toward advertising and its intended effects. Further research could investigate to what extent this pro-ad culture is stemming from children’s socialization at home, as their attitudes toward advertising are also significantly influenced byfamily background variables (e.g. education).It might be questioned, however, whether these attitudes constitute a genuine, unambiguous dimension of advertising literacy. On the one hand, negative attitudes (and avoidance tendencies) do seem to reflect some form ad skepticism or vigilance, and are found to be intrinsically connected to the dimensions of advertising literacy and related factors. On the other hand, these effects are rather small, and a number of interaction effects suggest that such a disposition can also imply a sort of ‘closedness’, impeding elaborate, critical reflection on advertising and its tactics.In other words, future research might consider to rethink advertising literacy as an ‘affective defense’.