Gender Stereotypes and Sexualization in Italian Children’s Television Advertisements

The objective of the present study was to examine gender stereotypes and sexualization in Italian television advertisements aimed at children. For this purpose, content analysis method was used to analyze 185 commercials broadcast from 6 pm to 8:30 pm over the three Italian television channels dedicated to children which attract the largest audience share. In order to allow comparison with previous research on gender stereotyping, two raters coded a series of variables common to similar investigations (i.e., verbs used in the ads, voice-over, voice-over message, setting, pace of the ads, activities performed by the main characters). Results indicate that gender stereotypes are still prevalent in Italian children’s television commercials. Furthermore, although the level of sexualization of children is low, we found that girls tend to be more sexualized than boys, especially in terms of “adultification.” Our findings provide an updated picture of children’s advertising in Italy by also expanding the literature on gender role stereotyping and sexualization in television commercials.


Introduction
The content of television advertisements has been an important object of study for many scholars. Commercials constitute indeed one of the main elements of the social pressure associated with stereotyped views of the world and society, especially with gender roles stereotypes (Allan & Coltrane, 1996;Valls-Fernández & Martínez-Vicente, 2007). It has been demonstrated that television influences children and adults quite intensely (Herrett-Skjellum & Allen, 1996;Ward & Grower, 2020;Wille et al., 2018). However, while basic foundations of gender have been laid in adults, children are still forming their values and beliefs. They are more vulnerable to many types of images or stereotypes presented to them, particularly those with audio and visual reinforcement (Davis, 2003). Television and television advertising can thus be considered a means by which children learn about gender behaviors and roles.
Researchers in communication and television commercials have frequently expressed concern at statistics which show fewer women in authority roles than men (Valls-Fernández & Martínez-Vicente, 2007), more women than men appearing in subservient and decorative roles, and men having both more important roles and a greater range of roles to fulfill (Valtorta et al., 2016). Although it would be difficult to argue that advertising's representations of women in subordinate or decorative roles are solely responsible for limiting the aspirations of young girls (see Yoder et al., 2008), researchers have demonstrated how advertising acts conservatively to reinforce outdated images (Aubrey & Harrison, 2004;Wille et al., 2018). In this regard, social cognitive theory indicated that children learn their gender roles through observation, identification, as well as through rewards and punishment (Bandura, 2009). By watching members of their own gender, identifying with them, and then imitating that behavior, children can discover how they are supposed to act and feel. Children learn a great deal from television about gender-typed behaviors because it provides them with many models readily available for observation and identification (Bandura, 2009;Ward & Grower, 2020).
Despite the potential influence of television advertising on children, few content analyses of children's commercials have been published, and none recently. Starting from these arguments, the purpose of the present study was to provide an updated picture of children's advertising in Italy by examining gender-role differentiation and sexualization, considered here as the imbuing of adult sexuality upon a child (see Zurbriggen et al., 2007).

Gender Stereotypes and Sexualization in Television Advertisements
Over the last thirty years, research has been particularly interested in the portrayal of men and women in television commercials (e.g., Furnham & Mak, 1999;Furnham & Paltzer, 2010; see Furnham & Lay, 2019 for a review). In one of the first content analyses of adult-oriented television advertising, McArthur and Resko (1975) found that women were depicted as younger than men, more likely to be married than men, and employed in traditionally female-dominated occupations. Women were also found predominantly in the home, and they were used less often as spokespersons. Crucially, when they were shown as product users, they did not play the role of authorities. Furnham and Mak (1999) reviewed and compared 14 studies conducted in five different countries over a period of 25 years. The results showed a common pattern, demonstrating the universality of gender role stereotyping in television commercials. A recent advertising analysis on the Ukrainian popular television channels indicated that women were mainly represented as housewives, nurses, or as decoration to power men (Kitsa & Mudra, 2020). In line with these results, by examining gender portrayals in television advertising in the Middle East (i.e., Bahrain, Kuwait, Oman, Qatar, Saudi Arabia, and the United Arab Emirates), Khalil and Dhanesh (2020) found that males were more likely to be depicted in non-familial roles than in familial roles, whether as fathers or husbands. Instead, females were more likely than men to be portrayed in domestic roles.
In the most recent study on the topic conducted in Italy (i.e., the context of the present research), Valtorta and colleagues (2016) analyzed 287 television commercials and found that women were portrayed as younger and more engaged as caregivers than men. Furthermore, men were more likely than women to be represented as independent, athletic, and competent. In the same cultural context, Zogmaister and Castelli (1998) analyzed 81 child-oriented advertisements and found significant differences in language and setting. In particular, whereas ads aimed at girls contained verbs that referred to positive social interactions, those aimed at boys were characterized by verbs related to competition. Furthermore, in line with several findings concerning adult-oriented commercials (e.g., Kitsa & Mudra, 2020;Valtorta et al., 2016), Zogmaister and Castelli (1998) found that young girls were predominantly portrayed at home and indoors, while young boys were more frequently depicted outdoors. Studies focusing on different cultural contexts led to similar conclusions. For example, a study conducted by Peirce (1989) looked at the young characters pictured on prime-time television programs and indicated that prime-time television's characters performed different activities depending on their gender. Girls who played with dolls and talked on the phone were likely to be at home. Instead, commercials showing boys participating in sports or acting as disc jockeys indicated more out-ofhome settings. In another analysis of children's programming, Barcus (1983) found that 29% of boys exhibited highly aggressive behaviors, compared to 16% of girls. Smith (1994) conducted a content analysis of gender differences in 218 children's advertising in the United States and found that advertisers placed boys in settings outside their homes more often than girls. Similarly, Browne (1998) examined gender role stereotyping in television commercials aimed at children in the United States and Australia by demonstrating that boys were depicted as being more knowledgeable, active, and aggressive than girls. Furthermore, non-verbal behaviors involving dominance and control were associated more with boys than girls. More recently, Furnham and Saar (2005) showed that young girls depicted in British and Polish children's advertisements were most likely to be portrayed in home settings than in any other environment.
As reported above, research on adult-and child-oriented advertising has revealed differences in gender representations. Overall, women and girls are more often shown as product users, in dependent roles, at home, and associated with domestic products or toys. Furthermore, research on nudity in television commercials has also highlighted gender differences. In particular, more men than women are fully dressed; conversely, more women than men are depicted undressed, exhibit more "sexiness," and are depicted as sexual objects (Matthes & Prieler, 2020;Valtorta et al., 2016).
Although advertisers are usually careful not to sexualize girls in a direct manner, several advertising techniques do so indirectly. Many scholars (e.g., Rodríguez et al., 2016;Zurbriggen et al., 2007) stated that children, and girls especially, are sometimes depicted as counterparts to adult versions. In their analysis of children and teenagers' commercials in magazines (e.g., Cosmopolitan) over a 40-year period, O'Donohue and colleagues (1997) found 38 ads (1.5%) that portrayed children and teenagers sexually. Furthermore, 85% of these ads focused on girls who often featured with sexualized adult women and were posed in matching clothing or seductive poses. In this framework, young girls were "adultified," and adult women were "youthified." More recently, Díaz-Bustamante-Ventisca and Llovet-Rodríguez (2017) examined the sexualization of girls spread through social networks and provided evidence of the negative correlation between sexualizing representations of girls and people's perceptions of girls' intellectual, social, and moral capacities. Likewise, by conducting a content analysis, Graff and colleagues (2013) showed that portrayals of girls in American teen magazines have become increasingly sexualized from 1994 to 2011. Crucially, there is extensive evidence that young people are avid consumers of these forms of media (Rideout et al., 2010;Slater & Tiggemann, 2016), making it likely that girls are exposed to a high degree of sexualized media messages from a young age. Despite the relevance of these considerations, it is worth mentioning that, as far as we know, no research has examined sexualization of children by specifically focusing on television advertising dedicated to them. The present study aimed to provide an updated picture of gender-stereotypical representations of girls and boys in Italian television commercials targeting children and to fill the gap in the literature on sexualization.

The Present Study
Television advertisements are one potential source of gender stereotypes for children. Despite the wide diversity of media available nowadays, television continues to be one of the most popular and widely used media among children (Rideout, 2015). Given that no content analysis of children's commercials has been published recently, the present research was designed to observe the differences between Italian television advertisements aimed at girls and aimed at boys in terms of the main variables used as stereotype indicators in prior advertising studies (i.e., verbs used in the commercials, voice-over, voice-over message, setting, pace of the advertisements, activities performed by the main characters; Zogmaister & Castelli, 1998). Furthermore, we intended to examine differences between the girls and boys portrayed in child-oriented commercials by analyzing the activities in which they were engaged within the advertisements and their level of sexualization. Zurbriggen and colleagues (2007) stated that sexualization occurs when a person's value mainly comes from their sexual appeal and appearance. Of relevance to the present research, the scholars described some of the most significant results that emerged in the literature on sexualization (e.g., Cook & Kaiser, 2004;Merskin, 2004) and indicated that advertising generally presents the public with images of women who are provocatively "dressed down" to look like little girls and young girls who are "dressed up" as grown women. According to Gerding Speno and Aubrey (2018), these two types of depictions can be conceptualized as part of the same system of messages known as "age compression," a practice in which women and girls in various media are compressed to an age in which they are most valued for their sexual function. One of the consequences of age compression is that kids are bombarded with adult content and messages that carry specific ideas about what it means to be an adult. Young girls receive messages about how to be sexy and attractive, and they appear in imagery complete with the trappings of woman's adulthood, such as mature clothing, jewelry, and makeup (O'Donohue et al., 1997). In the light of these considerations and previous literature on the topic (e.g., Gerding Speno & Aubrey, 2018;Loh, 2016), we decided to operationalize sexualization of children as "adultification," namely the inappropriate imposition of sexuality upon kids in terms of clothing, jewelry, fashion accessories, grown-up postures and gestures.
Based on the abovementioned findings, we formulated the following hypotheses: Hypothesis 1 commercials aimed at females would be more likely to contain prosocial (vs. antisocial) verbs. Instead, commercials aimed at males would be more likely to contain antisocial (vs. prosocial) verbs. Furthermore, we assumed that advertisements aimed at girls (vs. boys) would contain a higher mean number of prosocial verbs and that advertisements aimed at boys (vs. girls) would contain a higher mean number of antisocial verbs.
Hypothesis 2 female-oriented commercials would be more likely to use feminine (vs. masculine) voice-overs. Vice versa, male-oriented advertising would be more likely to use masculine (vs. feminine) voice-overs.
Hypothesis 3 commercials aimed at females would be more likely to convey warm or interaction-related (vs. technical or competitive-related) contents. Instead, commercials aimed at males would be more likely to convey technical information about the products or competitive-related (vs. warm or interaction-related) contents.
Hypothesis 4 female-oriented commercials would be more likely to take place in house or home settings (vs. outdoors). On the contrary, male-oriented commercials would be more likely to take place outside and in non-domestic environments (vs. indoors).
Hypothesis 5 male-oriented commercials would be more fast paced than femaleoriented commercials.
Hypothesis 6 both female and male characters would be engaged in stereotypical games or with stereotypical toys (e.g., girls with dolls and boys with brainteasers).
Hypothesis 7 girls would be more sexualized than boys in terms of "adultified" representations.

Sample
We recorded advertisements broadcasted by three Italian television channels dedicated to children between 4 and 14 years of age: Boing, Cartoonito, and K2. These three channels are broadcast nationwide and have the largest audience share (see https://www.auditel.it/dati/). Commercials were recorded in the week from 13 to 19 November 2019, during the peak audience timeframe (i.e., from 6 pm to 8:30 pm). Repeated advertisements, institutional advertisements, movie promotions, and the channels' self-promotions were omitted. The final sample was composed of 185 different commercials.

Variables and Coding
Two Italian-speaking raters (one female and one male) independently coded the characteristics of the advertisements and the central figures of each commercial. 1 Three different procedures were adopted to investigate gender stereotypes and sexualization of children. We first examined gender stereotypes by considering specific characteristics of the ads. Then, we investigated gender representations by focusing on the main figures and the activities in which they were engaged within the commercials. Finally, we explored the presence of sexualization by examining clothing, jewelry, fashion accessories, postures, and gestures of the children previously classified as main figures. The coding schemes adopted for gender stereotypes were adapted from Zogmaister and Castelli (1998). Regarding the characteristics used to examine sexualization, we adapted the items outlined by Graff et al. (2013) and Rodríguez et al. (2016).

Gender Stereotypes and Characteristics of the Ads
Each commercial was codified as aimed at females, males, or both. In order to operationalize whether the ads were aimed at females or males, it was defined that all the characters had to be females or males, respectively. Instead, commercials were classified as aimed at both if the characters were females and males together. Each advertisement was coded on the basis of the following variables: the verb type that our raters thought was the most used in the commercial (1 = prosocial; 2 = antisocial; 3 = neutral) and the numerosity of prosocial and antisocial verbs; the voice-over (1 = feminine; 2 = masculine; 3 = mixed); the content of the message conveyed by the voice-over, that is whether the message concerned emotional aspects and interpersonal warmth, technical information about the product, or both (1 = warm; 2 = technical; 3 = mixed); the setting (1 = indoor/domestic; 2 = outdoor/non-domestic; 3 = mixed); the pace of the advertisement, namely the speed of the commercial operationalized as the ratio of the number of frame changes to the total length of the ad. ANOVAs, chi-square tests, and Fischer's exact tests were then calculated for gender differences of the target of the ads within each coding category.

Gender Stereotypes and Performed Activities
The central figures of each advertisement were defined as the children involved in a verbal or non-verbal interaction that was the focus of the commercial. Thus, our raters first identified those children who were involved in the main interaction of the ad, then they proceeded with the coding. According to this procedure, in some cases, more than one child per advertising was coded. For each main figure, the activities performed were observed and coded as 1 = stereotypical, 2 = counter-stereotypical, or 3 = neutral. Stereotypical female activities included anything associated with sociability and self-care behaviors, such as playing with dolls or "house" and putting on makeup (see Olsson & Martiny, 2018). Vice versa, stereotypical male activities included anything related to smartness and competition, such as solving brainteasers and fighting (see Olsson & Martiny, 2018). Finally, neutral activities involved occupations that are not typically regarded as feminine or masculine, such as singing or listening to music. When a female character was engaged in a stereotypical female activity and a male character was engaged in a stereotypical male activity, the activities performed were coded as stereotypical. Conversely, when a female figure performed a stereotypical male activity and a male figure performed a stereotypical female game, the activities were coded as counter stereotypical. All the other activities were coded as neutral.

Sexualization and "adultification"
For each child defined as a central figure, grown-up/sexy clothing (e.g., shirt and tie for boys, evening dress for girls), jewelry (e.g., long necklaces, rings), fashion accessories (e.g., belts, bags, sunglasses), grown-up/sexy postures and gestures (e.g., boys or girls winking at the camera in an allusive way) were observed and coded as present or absent. Table 1 reports more details of the items used to investigate sexualization and "adultification."

Reliability
The two raters carefully coded the advertisement data based on the coding characteristics explained in the previous section. This process involves subjective judgment, but inter-rater reliability can help to achieve reliable and quantifiable results (Cre-

Items Clothing
Tight-fitting clothing (e.g., leggings) Very short clothing and/or items that leave the belly and/or the back exposed Grown-up clothing (e.g., trench coats, blazers) Imitation or adult jewelry (e.g., long necklaces) Adult accessories (e.g., belts, bags) Poses/Gestures Sexy poses/gestures Grown-up poses/gestures  swell, 2003). The level of inter-rater agreement was ascertained using Cohen's kappa test (1960). This statistic was designed to estimate the consensus between two raters. A value of 0 on kappa indicates that the two raters did not agree with each other any more than would be predicted by chance alone. Kappa values from 0.41 to 0.60 are considered moderate, and values above 0.60 as substantial (see Stemler, 2004). Our inter-rater reliability results revealed more than 97% agreement with regard to all the characteristics of the ads (k = 0.98 for the verb type; k = 0.97 for the numerosity of the verbs; k = 0.99 for the voice-over message; k = 0.97 for the setting; k = 1 for the voiceover and the pace of the advertisement), 100% agreement (k = 1) on the coding of the activities performed, and more than 88% agreement on the coding of sexualization and "adultification" (k = 0.88 for tight-fitting clothing; k = 0.97 for grown-up clothing and sexy poses/gestures; k = 0.98 for grown-up poses/gestures; k = 1 for very short clothing, jewelry, and accessories).

Gender Stereotypes and Characteristics of the ads
Of the 185 commercials analyzed, 36% (n = 67) were aimed at females, 22% (n = 41) were aimed at males, and 42% (n = 77) were aimed at both.

Verb type
Analyses of gender of the target and verb type indicate significant differences between ads aimed at females, males, and both as a function of the verbs used in the commercials, χ 2 = 65.54, df = 4, p < .001, Cramér's V = 0.42 (see Fig. 1). In line with our assumption (Hypothesis1), when the commercials were aimed at females, prosocial and neutral verbs were the most used (51% and 48%, respectively; n = 34 for prosocial verbs and n = 32 for neutral verbs). Antisocial verbs were adopted only in 1% of the cases (n = 1). Instead, when the ads were aimed at males, antisocial verbs were the most employed (59% of the cases; n = 24), followed by neutral and prosocial verbs (34% and 7%, respectively; n = 14 for neutral verbs and n = 3 for prosocial verbs). Finally, when the targets of the ads were both males and females, neutral verbs were employed more often (61% of the cases; n = 47) than prosocial and antisocial verbs (25% and 14%, respectively; n = 19 for prosocial verbs and n = 11 for antisocial verbs).

Numerosity of the Verbs
Levene's test was used to determine whether classic ANOVA was appropriate for this analysis. Results revealed a significant test for both prosocial, F(2,182) = 22.80, p < .001, and antisocial, F(2,182) = 87.53, p < .001, verbs. Thus, two one-way (target of the ads: females vs. males vs. both) Welch's ANOVAs were conducted to compare the mean number of prosocial and antisocial verbs among the commercials aimed at females, males, and both. The first Welch's ANOVA indicated a significant difference among advertisements aimed at females, males, and both in using prosocial verbs, F(2,100.60) = 11.05, p < .001, est. ω 2 = 0.10. In line with our assumption (Hypothe-sis1), post-hoc comparisons with the Games-Howell method showed a significantly higher mean number of prosocial verbs in ads aimed at females (M = 0.97, SD = 1.28) than males (M = 0.15, SD = 0.53), p < .001, and both (M = 0.26, SD = 0.52), p < .001. Furthermore, prosocial verbs were equally adopted in ads aimed at males and both, p = .507. The opposite trend emerged from the second Welch's ANOVA, which showed a significant difference among commercials aimed at females, males, and both in using antisocial verbs, F(2,73.73) = 22.80, p < .001, est. ω 2 = 0.19. More specifically, results revealed a greater mean number of antisocial verbs in ads aimed at males (M = 1.12, SD = 1.14) than females (M = 0.01, SD = 0.12), p < .001, and both (M = 0.18, SD = 0.48), p < .001. Furthermore, antisocial verbs were not equally used in ads aimed at females and both, p = .001.

Voice-over
Given that the cell regarding masculine voice-overs in commercials aimed at females was empty and some other statistics were low (i.e., mixed voices in ads targeting males and in ads targeting females, and feminine voices in commercials aimed at males), we performed Fisher's exact test. The existence of significant differences between commercials aimed at females, males, and both with respect to voice-over was confirmed, p < .001 (see Fig. 2). As assumed (Hypothesis2), when the ads were aimed at females, feminine voice-overs were the most common (99% of the cases; n = 66), whereas masculine voices were never used. Mixed voices were adopted only in 1% of the cases (n = 1). When the commercials were aimed at males, masculine voice-overs were the most employed (96% of the cases; n = 39), followed by feminine and mixed voices (2% for both the categories; n = 1). Finally, when the targets of the advertisements were both females and males, masculine voices were the most used (53% of the cases; n = 41), followed by feminine and mixed voices (33% and 14%, respectively; n = 25 for feminine voices and n = 11 for mixed voices).

Voice-over Message
Given that the cell regarding warm messages in ads targeting males was empty, we conducted Fisher's exact test. We found significant differences according to the gender of the target in the voice-over message, p < .001 (see Fig. 3). In line with our expectation (Hypothesis3), when the advertisements were aimed at females, the voice-over messages conveyed mostly warm and mixed contents (48% and 42%, respectively; n = 32 for warm contents and n = 28 for mixed contents). In 10% of the cases (n = 7), the messages communicated technical information. Vice versa, when the commercials were aimed at males, the voice-over messages were especially technical (95% of the cases; n = 39). In 5% of the cases (n = 2), the messages were mixed. Crucially, warm messages were never conveyed. When the targets of the ads were both females and males, the messages were mostly technical (65% of the cases; n = 50). In 25% of the cases (n = 19), the messages were mixed, and in 10% of the ads (n = 8), the content of the voice-over communication was warm.

Setting
As for the setting where the commercials took place, we found significant differences with regard to the gender of the target, χ 2 = 30.82, df = 4, p < .001, Cramér's V = 0.29 (see Fig. 4). As expected (Hypothesis4), when the ads were aimed at females, the main characters (i.e., girls) appeared more frequently in house or home settings (73% of the cases; n = 49) than outside or in non-domestic contexts and in mixed settings (21% and 6%, respectively; n = 14 for outside or in non-domestic contexts and n = 4 for mixed settings). A similar result emerged when the commercials were aimed at both males and females. Crucially, when the advertisements were aimed at males, the main characters (i.e., boys) appeared more frequently outside and in non-domestic settings, such as the street (54% of the cases; n = 22), rather than in domestic or mixed contexts (29% and 17%, respectively; n = 12 for domestic contexts and n = 7 for mixed contexts).

Pace of the Advertisement
The two raters independently counted the number of frame changes for each commercial. Then, we computed the pace of each advertisement by dividing that value by the total length of the ad. 2 Levene's test was not significant, F(2,182) = 2.19, p = .115. Thus, a one-way (target of the ads: females vs. males vs. both) ANOVA was conducted to compare the mean pace among the commercials aimed at females, males, and both. The analysis revealed a significant difference among commercials aimed at females, males, and both in the pace of the ads, F(2,182) = 5.05, p = .007, η p 2 = 0.05. As assumed (Hypothesis5), results showed a greater mean pace in commercials aimed at males (M = 0.80, SD = 0.16) than in those aimed at females (M = 0.69, SD = 0.14), p = .006, but not than both (M = 0.73, SD = 0.20), p = .074. Furthermore, the pace of the ads was similar in advertisements aimed at females and both, p = .786.

Gender Stereotypes and Performed Activities
Of the 289 coded children, 59% (n = 171) were females and 41% (n = 118) were males. It is worth mentioning that only 17% (n = 48) of children appeared in male-oriented commercials, while 36% (n = 104) of them appeared in female-oriented advertisements. The remaining characters (47%, n = 137) appeared in commercials aimed at both males and females. The analysis of gender and activities performed produced non-significant differences, χ 2 = 3.45, df = 2, p = .178, Cramér's V = 0.11. In line with our assumption (Hypothesis6), both females and males were represented in a similar way, namely while they were performing their respective stereotypical activities. As shown in Fig. 5, more than half of females and males (66% and 56%, respectively; n = 113 for females and n = 66 for males) were depicted as engaged in stereotypical games. Only 8% of both (n = 14 for females and n = 10 for males) appeared instead where counter-stereotypical tasks were performed.

Sexualization and "adultification"
Given that most of the indicators were coded as absent by our raters, we merged some of them together. In particular, we combined the indicators concerning clothing, accessories, and jewelry to create a single index of physical appearance. Instead, we merged the two indicators about poses and gestures to create a single index of movement. We considered sexual/grown-up appearance and sexual/grown-up movements as present when, for each central figure, at least one of the indicators (i.e., tight-fitting clothing, very short clothing, grown-up clothing, imitation or adult jewelry, and adult accessories for the index of physical appearance; sexy poses/gestures and grown-up poses/gestures for the index of movement) was coded as present by the two raters. As reported in Table 2, although the level of sexualization of children was low, most of the few cases in which our items were coded as present involved a female character, by thus confirming that this phenomenon mainly affects the female gender also among children (Hypothesis7).

Discussion
The results of the present study make it apparent that gender stereotypes are still prevalent in Italian children's television commercials. In particular, we found that advertisements aimed at girls contained more prosocial verbs and feminine voiceovers, were more likely to take place in house or home settings, and were more likely to convey warm or interaction-related contents. On the contrary, male-oriented commercials contained more antisocial verbs and masculine voice-overs, were more likely to take place outside and in non-domestic environments, and were more likely to convey technical information about the products or competitive-related contents. The activities performed by the characters confirmed this gender-role differentiation. While girls were more likely to play with dolls or "house" and put on makeup, boys were more likely to run, fight, and solve brainteasers. Furthermore, we found that most of the few cases in which our raters identified sexualized features involved a female rather than a male character.
Our results also indicated that advertisements aimed at boys were more fast paced than those aimed at girls by indirectly confirming the idea that males are stereotypically more agentic than females, namely more active and engaged in diverse activities. As indicated in previous literature on the topic (e.g., Welch et al., 1979;Zogmaister & Castelli, 1998), using a large number of frame changes can be considered a technical tool designed to enhance the action, speed, and toughness of masculine toys, and may, in turn, reflect prescriptive gender stereotypes about what boys and girls should do. This difference in the pace of the advertisements is particularly relevant not only for the literature on stereotypes but also for development psychology. Indeed, some studies (e.g., Christakis, 2009;Nigg, 2006;Sigman, 2007) have demonstrated the detrimental effects of fast-paced and violent media on the development of attention-deficit/hyperactivity disorder (ADHD) or ADHD-related behaviors (i.e., attention problems, hyperactivity, and impulsivity) among children. According to the fast-pace-arousal-habituation hypothesis, the fast pace of entertainment media may increase arousal by triggering repeated attention shifts in the user (Lang et al., 2000). After frequent exposure to fast-paced media, children might get habituated to this arousal lift, thereby decreasing their baseline arousal level. Low baseline arousal may then cause ADHD-related behaviors (see Nikkelen et al., 2014). Even if more research is needed to deeper investigate this specific association, our results seem to be potentially relevant to other domains than social psychology by thus serving as an important starting point for future studies.
It is interesting to compare the present findings with those of similar investigations. For example, our study lends support to Davis' (2003) results, as the location of the commercials played a key role in determining the gender of the characters of the ads. Being outdoors rather than indoors or in domestic environments dramatically increased the probability of a character being male. This finding is consistent with previous studies that highlighted that commercials targeted at girls took place indoors (e.g., in a girl's bedroom), whereas male-oriented advertising took place outdoors (Merskin, 2002;Seiter, 1993). This suggests the appropriate place for girls is still in the home, while boys are afforded access to privileged outdoor settings.
Our results are also consistent with those of Owen and Padron (2015). They analyzed the narratives of contemporary action figure toys and found that narratives of girls' action figures in comparison to boys' action figures contained more social words. Instead, the narratives of boys' action figures contained more adjectival references to power and destructive action. Together with our findings on verb type and voice-over contents, these data reflect stereotyped gender roles, with linguistic associations to masculine attributes of aggression and traditional feminine attributes of social relatedness. Language has a powerful effect on perceptions and behaviors. When the language used to promote children's toys is gendered, a message that traditional stereotypes are appropriate is conveyed. An important issue regards the effect of these messages on children. A large body of research demonstrates indeed that media representations of stereotypical gender roles affect children's perceptions of gender (Henning et al., 2009;Pike & Jennings, 2005;Weisgram et al., 2014). Further, acceptance of gender-stereotyped beliefs in children has been associated to bullying and ostracism behaviors directed toward gender-nonconforming children (Meyer, 2009). Crucially, media reinforcement of stereotyped gender roles imparts lessons about what roles are allowed and which are not as based on gender. Boys, overwhelmed with violence and dominance in media messages, are not afforded socialization experiences that teach cooperation. Vice versa, girls, limited by messages promoting only care for others and passivity, are deprived of lessons endorsing power and self-agency.
Comparing our results to those found in past research conducted in Italy (i.e., Zogmaister & Castelli, 1998), we found a continuation of gender stereotyping of characters in commercials. In particular, we considered the findings from Zogmaister and Castelli (1998) concerning the variables we measured in a very similar way (i.e., verb type, voice-over, and setting) and observed that little has changed in twenty years. Girls are still shown performing their activities in house or home settings, while boys are shown doing exciting things in the outdoor world. In 1998, Italian female-oriented commercials contained more feminine voices and interaction-related contents, while commercials targeting males contained more masculine voice-overs and antisocial verbs, as in our study (for more details, see the Supplementary Material on OSF). 3 Overall, these findings are similar to those found in other research on adult-oriented advertisements performed not only in Italy (Valtorta et al., 2016) but also in the United States (Coltrane & Adams, 1997), Spain (Valls-Fernández & Martínez-Vicente, 2007), Portugal (Neto & Pinto, 1998), Ukraine (Kitsa & Mudra, 2020), Turkey (Uray & Burnaz, 2003), Great Britain and Poland (Furnham & Saar, 2005). This consistent pattern is particularly alarming. Although television viewers often assert that commercials do not affect them negatively (Gunther & Thorson, 1992), studies have shown that heavy television viewing may influence people's (especially children's) perceptions of behaviors associated with gender (Ward & Grower, 2020). It may not be advertisers' responsibility to reflect precise images of our society. However, the burden of responsibility is on the advertisers when they fail to reflect the changes in such stereotypes in society.
Through this research, we provided the most recent data about gender stereotypes in Italian children's television commercials. Furthermore, we documented the presence of sexualization. Although this representation was not particularly salient, we found that females tended to be more sexualized than males, especially in terms of "adultification." In particular, girls were more often depicted while making grownup poses, and wearing adult accessories and tight-fitting clothing. A large body of research has offered evidence of negative consequences for girls when they are sexualized or exposed to sexualized representations (see Zurbriggen et al., 2007). First, there is indication that girls exposed to sexualizing content are more likely to experience body dissatisfaction (Groesz et al., 2002;Holmstrom, 2004), depression (Durkin & Paxton, 2002), and lower self-esteem (Rivadeneyra et al., 2007). Furthermore, girls' relationships with other girls might be affected by these representations, as such relationships may become dangerous grounds where girls reject other girls for reasons having to do with conformity to a beauty ideal (Zurbriggen & Morgan, 2006). In addition, girls' relationships with boys and men might be impacted in that exposure to sexualized media has been shown to relate to children's views on dating, sexual harassment, and attitudes toward sexual violence (Fox et al., 2015;Ward & Friedman, 2006).

Limitations and Future Directions
Despite the relevance of our research and findings, some limitations should be considered. Because we analyzed only commercials from three television channels in the peak audience timeframe, we might not be fully able to reflect the pool of television ads in other time periods and channels. Future studies should expand our results by considering larger and more diverse samples. Research should also focus on other kinds of media potentially used by children, such as YouTube or other social network sites. For example, a recent investigation explored the visual elements of gender stereotypes in children's toy advertisements on Facebook by showing that toys targeted at boys and girls are stereotypically promoted (Azmi et al., 2021).
In addition, for gender representations and sexualization, we investigated only primary characters, a limitation that restrains the amount of information gleaned from the television advertisements to a certain extent. Future research should consider including more characters.
Finally, the present study cannot demonstrate that negative social consequences arise from the representations examined -and the same is likewise true of every content analysis study -, yet the possibility of negative social consequences does exist. Future studies might investigate the success or failure of using stereotypes in children's advertising.

Conclusions
Our study was able to show that gender stereotypes and traditional gender roles are still present in Italian television advertising aimed at children. As reported by several authors (e.g., Bandura, 2009;Mathur & Moschis, 1999), the communication media are socialization agents that continuously affect individual's behaviors and attitudes. As a significant media component, advertising also wields an influence within the socialization process. It is considered to have unintended consequences on the values and lifestyles adopted by members of society, especially children (Pollay, 1986;Wille et al., 2018). We hope our research and findings help to spur a discussion among scholars and advertisers on the global dominance of gender stereotyping in advertising, in order to promote a more gender-equal approach toward gender roles.

Notes
1. The two raters were undergraduates from our Psychology department. They did not know the main aim of the study, and a careful explanation was given about the coding schemes before rating. At the end of the task, the raters were debriefed by the project team to receive feedback about the coding scheme and discuss the results and any doubts that might have been raised during the coding task. After the debriefing phase, given the central role the two raters had in this project, we decided to include them among the co-authors of the current study. 2. We checked whether the length was different as a function of the target of the advertisements. Levene's test was significant, F(2,182) = 4.41, p = .013. Thus, we conducted Welch's ANOVA. Results revealed a significant effect of target, F(2,112.95) = 4.48, p = .013, est. ω 2 = 0.04. However, post-hoc comparisons with the Games-Howell method indicated that the ads aimed at males (M = 15.73, SD = 4.55) and females (M = 16.70, SD = 5.20) did not significantly differ in their length, p = .568. Similarly, the length of the ads aimed at females did not significantly differ from those aimed at both (M = 19.03, SD = 7.48), p = .077. The only significant comparison was between the ads targeting males and those targeting both, p = .010. We also conducted additional analyses to verify the potentially relevant role of the length of the ads. In particular, we performed Pearson correlations between the length score and the amount of antisocial and prosocial verbs identified by our raters. Results indicated a non-significant association between the duration of the commercials and the number of antisocial (r = .01, p = .873) and prosocial (r = .05, p = .489) verbs. In the light of all these non-significant results and considering that we did not have specific assumptions about this variable, we decided to not consider the length of the commercials in the main analyses. 3. All data, R syntax for the analyses, and Supplementary Material are available through the Open Science Framework: https://osf.io/2kzhc/.