Introduction

This study is concerned with the representation of men and women in TV advertisements in the United Kingdom and Saudi Arabia. It uses a content analysis methodology to assess samples of 164 televised advertisements from each country to find out whether each gender is utilised in different ways and in relation to different products. This research follows a long line of studies (e.g., Furnham and Bitar 1993; Furnham and Skae 1997; Livingstone and Green 1986; Manstead and McCulloch 1981) of gender representation in televised advertising in the UK, and other studies that have compared gender representations in British televised advertising with those in advertising from other countries (Furnham and Farragher 2000; Furnham and Mak 1999; Furnham et al. 2000). It also adds to earlier literature concerned with gender representation in televised advertising published in this journal (Downs and Harrison 1985; Mazzella et al. 1992; Neto and Pinto 1998), including some cross-cultural comparisons studies published in this journal (e.g., Bresnahan et al. 2001; Furnham and Mak 1999).

This study was the first attempt to compare gender representation in televised advertising between the UK and a Middle Eastern nation. Saudi Arabia differs in significant ways from the UK politically, socially and culturally. It operates under strict Islamic law, does not hold open elections, and is characterised by a society in which men and women have different rights. The media systems in the two countries also differ in significant ways (see Al-Saud 1997; Smith 1995). The media in Saudi Arabia are much more tightly controlled by government than is the case in the UK (Marghalani et al. 1998). Appearances on television by women, including in advertising, are governed by Islamic law which, in particular, prohibits certain forms of dress for women or girls from the age of six (Al-Makaty et al. 1996).

It is important to understand these differences between these two countries, particularly as they pertain to the effectiveness of advertising in their respective markets for several reasons. There is growing trade between Saudi Arabia and the United Kingdom and many British consumer products and services are found in the Middle East (Mooij 1998). In fact, expenditure on local and international advertising in Saudi Arabia is among the fastest growing of any national economy in the world (Pan Arab Research Centre Report 2001). Religiosity can be among the most influential cultural variables associated with consumer behaviour (Delener 1994). In Saudi Arabia, religious values have previously been identified as delaying the introduction of advertising into the major mass media and, in particular, on television (Al-Makaty et al. 1996). Further, the advertising systems in the two countries have different histories and different codes of practice that continue to place distinct restrictions on where advertising can appear and the way it can appear (Al-Yusuf 1989; Pan Arab Research Centre Report 2001; Usunier 1996).

This study provided an opportunity to investigate differences between a developed Western market and a developing Eastern market in the way advertisers promoted similar types of products and, of course, to compare, in each case, differences in the representations of women and men in advertising. In so doing, it will address issues raised above that concern whether televised advertising in the United Kingdom and Saudi Arabia display manifest differences in content and format that reflect political, religious and other cultural differences between the two countries. In placing an emphasis on gender-role portrayals, it will also extend a growing body of research literature, much of which has been published previously in Sex Roles, and provide a rare comparison of this kind between a western nation and a rapidly developing economy in the Middle East.

Previous research has emphasised pronounced attribute and role stereotyping in advertising associated with gender that has, according to some observers, placed women in more dependent and subservient and men in more authoritative and controlling positions (see Gunter 1995; Manstead and McCulloch 1981). Such stereotyping, in turn, often restricted the types of products women were used to advertise (Durkin 1985a, b; Thoveron 1987). Women were often depicted as little more than decorative features (Venkatesan and Losco 1975) or used to demonstrate products associated with traditional domestic roles or personal appearance (Dispenza 1975; Thoveron 1987). Portrayals that ran counter to these stereotypes were rare (Durkin 1985c). Later research from the United States has indicated, however, that television advertising has exhibited a shift in the nature and distribution of female and male role portrayals. Women are used much more often in authoritative user roles and promote a wider range of product types than had been the case 20 to 30 years earlier (Bartsch et al. 2000).

On the basis of the above evidence, one aim was to find out to what extent similar patterns of gender stereotyping occurred in UK and in Saudi Arabian television advertising. If the current findings were consistent with historical evidence, then some degree of gender stereotyping would be manifest in contemporary advertising. If trends observed in the United States had travelled to the United Kingdom, however, there might be some reduction in stereotyping of women in advertisements compared with that observed in studies conducted 10 or 20 years earlier. Given the traditional and religiously conservative nature of Saudi Arabian society, with concomitant restrictions on televised advertising (and especially with respect to the way women may appear; Al-Makaty et al. 1996) gender-role stereotyping was expected to be more pronounced in Saudi advertising than in British advertising.

Creation of an effective advertising campaign to target consumers from different cultural backgrounds can pose a serious problem for advertisers (Zhang and Harwood 2004). In global campaigns, or even national campaigns that occur in multi-cultural markets, it is important that advertisers avoid offending one community while trying to please another (Cheng and Schweitzer 1996; Pollay 1983). We investigated the content and presentation styles of televised advertising in two distinct national and cultural markets—Britain and Saudi Arabia. Special reference is made here to gender representation in advertising.

Britain is a mature market, whereas Saudi Arabia is a developing nation in which advertising has come to play an increasingly prominent part in the promotion of brands. Britain is home to some of the world’s leading advertising agencies and a major source of global advertising (Zenith Optimedia 2001). Saudi Arabia is an emerging target market for global advertisers and, as noted earlier, has witnessed one of the biggest growths of any country in advertising expenditure during the 1980s and 1990s (Al-Yusuf 1989). To ensure campaign success in a market as distinctive as that of Saudi Arabia, it is important for advertisers to be aware of the differences between this Islamic market and Western markets in relation to what is deemed to be acceptable or not in advertising (Srikandath 1991). An examination of gender representation in televised advertising provides one manifestation of differences between Middle East and West commercial markets and how to reach them.

Gender role stereotyping has been observed over many years to be a routine feature of television advertising. From the earliest studies conducted in the United States on this subject in the 1970s (e.g., Dominick and Rauch 1972; Courtney and Whipple 1974; McArthur and Resko 1975; O’Donnell and O’Donnell 1978) to research conducted across Western and Eastern nations at the turn of the millennium (Bartsch et al. 2000—USA; Bresnahan et al. 2001—Japan, Malaysia, Taiwan, USA; Furnham et al. 2000—Hong Kong, Indonesia, UK), results have consistently indicated that men and women are used differently in television advertisements.

Early research indicated that men were physically present more than were women especially in lead roles (Furnham and Bitar 1993—UK; Furnham and Farragher 2000—UK; Livingstone and Green 1986—UK; McArthur and Resko 1975—USA; Manstead and McCulloch 1981—UK; Neto and Pinto 1998—Portugal). It is the nature of the traits displayed and the types of roles occupied by lead characters of each gender that have given rise to the most concern. Indeed, the disproportionate presence of men and women was mostly confined to prime-time evening broadcasts, whereas during the daytime, women and men had an equal presence in advertisements (Culley and Bennett 1976—USA; Downs and Harrison 1985—USA).

In terms of traits, women are portrayed as dependent and unintelligent consumers, whereas men appear as experts and key information sources (Furnham and Skae 1997—UK). The authority of men in advertising has also been manifest in the tendency for most voice-overs in advertisements to be male rather than female (Furnham and Bitar 1993—UK; Furnham and Mak 1999—UK; Lovdal 1989—USA; Maracek et al. 1978—USA; O’Donnell and O’Donnell 1978—USA).

As for the nature of gender representation in advertisements, researchers have found that women are less likely to be shown in professional or occupation roles and more likely to be shown in domestic roles than are men (Furnham and Bitar 1993 - UK; Livingstone and Green 1986—UK; McArthur and Resko 1975—USA; Schenider and Schneider 1979—USA).

The settings in which men and women appear tend also to reinforce role bias between genders; women traditionally appear most often in domestic settings and men in work settings. Women are more often shown indoors and men outdoors (Bresnahan et al. 2001—Japan, Malaysia, Taiwan, USA; Ferrante et al. 1988—USA; Furnham and Schofeld 1986—UK; Lovdal 1989—USA).

Concern about gender stereotyping in advertisements has centred on the potential impact exposure to such depictions could have on consumers—especially children—in shaping public perceptions of gendered traits and gender roles. One argument is that repeated exposure to such content can cultivate distorted impressions of women through the suggestion that some social (e.g., domestic versus occupational) roles are appropriate only for one gender (Ferrante et al. 1988; Lovdal 1989). Empirical evidence has emerged to substantiate the hypothesis that gender representation in the media can affect people’s gender-related attitudes and behaviours (Gunter 1995; Mackay and Covell 1997).

Much of the research on gender representation in television advertising has been carried out in the United States. A growing body of work from other countries, however, has largely replicated American findings (Furnham and Mak 1999). Such studies have been conducted in Australia (Mazzella et al. 1992), Britain (Manstead and McCulloch 1981; Furnham and Schofeld 1986; Furnham and Bitar 1993; Livingstone and Green 1986), Canada (Rak and McMullen 1987), Japan (Arima 2003), Mexico (Gilly 1988), Italy (Furnham and Voli 1989), Korea (Kim and Lowry 2005), Portugal (Neto and Pinto 1998), and Turkey (Uray and Burnaz 2003).

In addition to research in Western industrialised societies, research has been undertaken in developing nations. Once again, gender role and trait stereotypes emerged that were consistent with those observed in North America and Europe in advertising in Africa (Mwangi 1996), Hong Kong, Indonesia (Furnham et al. 2000), Japan, Taiwan, and Malaysia (Bresnahan et al. 2001).

In Saudi Arabia, few previous researchers have analysed the content of television advertising. Razzouk and Al-Khatib (1993) analysed 279 television advertisements in Saudi Arabia between 1986 and 1988 and found that half (50%) featured Arab lead characters, and one in seven (14%) featured women, invariably in household, domestic roles. Al-Sharif (1994) analysed the appearance of children in Saudi Arabian television advertisements and focused mainly on the use of different Arab dialects in commercial messages.

The current study is a cross-national comparison of gender representation in televised advertising between a western, industrialised country (United Kingdom) and a developing nation in the Middle East with a distinct cultural and religious ethos (Saudi Arabia). As such, it represents one of the few studies carried out on the content of televised advertising in the developing world (Furnham et al. 2000; Mwangi 1996). Furthermore, it represents one of the few television advertising content studies conducted in a Muslim nation (see also Bresnahan et al. 2001; Furnham et al. 2000).

A study of television advertisements in Indonesia, for example, reported that men outnumbered women as lead figures and as voice-overs, but women were visually present in any roles in advertisements proportionately more than men were. There was clear occupational–domestic role stereotyping, with women more likely to be featured in domestic, home-based roles and men in occupational, out-of-the-home roles (Furnham et al. 2000).

On the basis of evidence produced by earlier studies of gender stereotyping in televised advertisements in the UK and other countries, the following hypotheses were investigated.

  1. H1:

    Male leads will outnumber female leads in television advertisements in both Britain and Saudi Arabia (see Furnham and Bitar 1993; Livingstone and Green 1986; Manstead and McCulloch 1981).

  2. H2:

    Male leads in television advertisements will appear in non-domestic roles more often than will female leads in both countries (see Furnham and Bitar 1993; Livingstone and Green 1986).

  3. H3:

    Female leads in television advertisements will appear in domestic roles more often than do male leads in both countries (see Furnham and Bitar 1993; Livingstone and Green 1986; Schenider and Schneider 1979).

  4. H4:

    Female leads will, on average, be younger than will male leads in television advertisements in both countries (Gilly 1988).

  5. H5:

    Female leads will appear in home-based settings more often than will male leads in television advertisements in both countries (Livingstone and Green 1986; Manstead and McCulloch 1981).

  6. H6:

    Male voice-overs will be more prevalent in television advertisements than will female voice-overs both countries (see Furnham and Bitar 1993; Furnham and Mak 1999; Lovdal 1989; Maracek et al. 1978).

  7. H7:

    Women will appear as leads more often than men will in advertisements for household and personal hygiene products in both countries (Thoveron 1987).

  8. H8:

    Men will appear as leads more often than women will in advertisements for non-domestic/household products in both countries (Furnham et al. 2000).

  9. H9:

    Overall, given the restrictions placed on the representation of women under Islamic law, gender stereotyping in televised advertisements will be more pronounced in Saudi Arabia than in the UK.

Method

Sample of Advertisements

Samples of television advertisements were video-recorded from Channel One of Saudi television and ITV1 (Independent Television) in Britain. Each is the most watched commercial television channel and regularly achieves the largest audience shares of any commercial channel in their respective countries. The Saudi advertisements were recorded over a 6-month period (May to October, 2000) between 9 and 11 p.m., Monday to Sunday evenings. The periods selected for sampling avoided major religious events that could significantly distort the regular television schedules, including numbers of advertising breaks shown. A total of 425 commercial message spots were recorded and yielded 164 distinct brand advertisements, which exhausted the brands advertised during this period.

The British advertisements were video-recorded between August 2000 and April 2001, between 9 and 11 p.m., Monday to Sunday evenings. These dates covered a period from the launch of a new television season, but excluded the month of December when the Christmas advertising campaign significantly changed the normal profile of brands advertised. Advertisements were more numerous on British television than on Saudi television, but many individual brand advertising messages are repeated more than once during a single evening. Video-recording stopped during the sampling period once 164 distinct advertisements had been captured to match the size of the Saudi sample.

Gender-related analysis was restricted to advertisements that featured characters aged 11 or over and in which the on-screen actors were not animated or cartoon characters. A total of 108 Saudi advertisements and 132 British advertisements satisfied these conditions. The age range in gender coding included younger actors than some earlier studies (Bartsch et al. 2000; Ferrante et al. 1988; Furnham and Skae 1997), and this age split is important in Saudi advertising because the Islamic dress requirements for women apply more strictly from the early adolescent years.

Coding Procedure

Content analysis was used to examine the advertisements (Krippendorf 1980). The coding framework was derived from a composite of earlier studies (Bartsch et al. 2000; Ferrante et al. 1988; Furnham and Bitar 1993; Lovdal 1989; Manstead and McCulloch 1981; McArthur and Resko 1975). The current analysis focused on ‘lead’ characters only and these represented one key unit of analysis. In a further analysis, voice-overs were also coded and were treated as a distinct set of units of analysis.

The following system was used to determine the lead characters in advertisements: (1) those who provide any information cues about the advertised product or service, or (2) those who use or hold the product, or (3) those who speak, or (4) those who appear longest. According to this system, the person(s) providing any information about the advertised product or service has/have the priority to be considered as lead figures. If none of the characters provided any product-related information, the second decision point would refer to the person using or holding the product, and then to the person who speaks, and finally to the person who appears longest.

Any individual character portrayed visually in a leading role in an advertisement was categorised as a lead figure. Each lead character and each voice-over was coded for gender (male, female, don’t know).

For each lead character, a number of further characteristics were coded. Following previous research (Ferrante et al. 1988; Livingstone and Green 1986), lead characters were categorised in terms of the roles they played in advertisements: (1) office work; (2) professional work; (3) household chores; (4) leisure; and (5) other.

Lead characters were categorised according to age. Coders were required to estimate to which of a number of age-groups the characters belonged: (1) new born to 10 years; (2) 11 to 19 years; (3) 20 to 29 years; (4) 30 to 39 years; (5) 40 to 49 years; and (6) 50 years and older.

Setting is a widely used category in advertising and gender representation research (see Bresnahan et al. 2001; Ferrante et al. 1988; Furnham and Schofeld 1986; Lovdal 1989; McArthur and Resko 1975), and it was coded here in relation to 11 indoor and 10 outdoor locations. The indoor settings included: living room, kitchen, bedroom, toilet, hallway, stairs, office, car showroom, restaurant, shop, and other. Outdoor settings included: street, sea or river, garden, desert, beach, field, mountains, forest, playground, and other.

Lead characters were also classified in terms of the product categories they advertised. Pilot research and other published research (Bresnahan et al. 2001; Livingstone and Green 1986; Lovdal 1989; McArthur and Resko 1975) was used to develop a taxonomy of 14 product categories: (1) food; (2) beverages; (3) automobiles; (4) services; (5) personal care products; (6) households cleaning products; (7) household appliances; (8) entertainment; (9) drugs and medicine; (10) retail stores; (11) restaurant; (12) dairy (milk, cheese, other milk-based products); (13) engine oil; and (14) other.

Coding Reliability

All advertisements were coded independently by two coders. The first author coded all advertisements, and separate British and Saudi coders were recruited to code subsets of the British and Saudi samples respectively. Each coder was provided with a videotape that contained sub-samples of 20 advertisements that were randomly selected from the Saudi and British samples for double-coding. Coders were provided with extensive training beforehand, which included practice runs using the coding frame on advertisements that did not form part of the main national samples. Coders met to discuss and resolve coding discrepancies after they completed their independent coding sessions. Inter-coder reliabilities, following the procedure outlined by Holsti (1969) were 90% for the Saudi advertisements and 99% for the British advertisements.

Results

Chi-square analyses were computed throughout on distributions of males and females in respect of specific attributes and roles within advertisements to test the significance of research hypotheses.

Male lead characters outnumbered female lead characters across both the Saudi and British samples. Among the Saudi advertisements analysed, more than one-half (53%) had male leads and less than one-half (47%) had female leads. Among the British advertisements, the distribution of male (52%) and female (48%) leads was similar. Overall these differences were not significant, X 2 – 0.63, df = 1, p < 0.47, which led to the rejection of H1.

Character Roles

Lead characters were classified in terms of whether they performed tasks associated with work, domestic chores, or leisure activities. In Saudi advertisements, male lead characters occurred in leisure roles (64% of male leads in leisure advertisements) significantly more often than did female leads 36%) X 2 = 6.05, df = 1, p < 0.02. Female leads (88% of leads) performed household roles significantly more often than did male leads (12% of leads) X 2 = 16.03, df = 1, p < 0.0001. Among British advertisements, there were no significant differences in the roles occupied by male and female lead characters. Hence H2 and H3 were supported by the findings from the Saudi advertisements but not by the findings from the British advertisements.

Male leads were significantly less likely in British advertisements (61% of leads) than in Saudi advertisements (78% of leads) to be depicted in professional roles, X 2 = 8.74, df = 1, p < 0.003. In addition, male leads in Saudi advertisements (64%) were more likely than their British counterparts (51%) to be featured in leisure roles, X 2 = 5.06, df = 1, p < 0.03. Among female leads, those in British advertisements (39%) were significantly more likely to be shown in occupational and professional roles than those in Saudi advertisements (22%) X 2 = 8.86, df = 1, p < 0.003.

Age

Most lead characters in both Saudi and British advertisements were assigned to two age groups: 20 to 29 and 30 to 39 years. In Saudi advertisements, the age-group distribution differed by gender, X 2 = 13.8, df = 4, p < 0.01. There were more female (63%) than male leads (35%) in the 20 to 29 range, and more male (28%) leads than female leads age 40 and older (10%). In British advertisements, lead characters were distributed across a wider age range than was found in Saudi advertisements, but there was no significant difference in the age profiles of male and female characters, X 2 = 4.1, df = 4, p < 0.1. The results for Saudi advertisements support H4, whereas those for British advertisements do not.

Comparing between Saudi and British advertisements, there was no overall difference in the age profile of male leads in each sample, X 2 = 2.2, df = 4, p < 0.2), but there was a significant cross-national difference for female leads, X 2 = 1.2, df = 4, p < 0.05. That is, there was a significantly higher proportion of female leads aged 20 to 29 in Saudi advertisements (63%) than in British advertisements (39%) X 2 = 7.62, df = 1, p < 0.006, and significantly more female leads aged 40 to 49 in British (22%) than in Saudi advertisements (5%), X 2 = 7.87, df = 1, p < 0.005.

Setting

Four principal settings emerged across Saudi and British advertisements: home, occupational, shop/retail, and outdoors. In Saudi advertisements, 41 female lead characters (69% of all female leads) appeared in home-based settings compared to 19 male lead characters (32% of all male leads), X 2 = 8.1, df = 1, p < 0.007. In contrast, a significantly greater number of male lead figures (32 or 78%) than of female leads (9 or 22%) appeared in outdoor settings, X 2 = 13.0, df = 1, p < 0.001, which supports H5. In British advertisements, there were no significant differences between male and female lead characters in their distribution across domestic, workplace and other outdoor settings, which lends no support to H5.

Cross-national comparisons revealed no significant differences between Saudi and British advertisements in the distribution of male leads across different settings. However, a significantly greater number of Saudi television advertisements featured female lead figures in home settings (68%) than did British advertisements (37%), X 2 = 12.2, df = 1, p < 0.001.

Voice-Overs

A total of 124 Saudi television advertisements featured a male voice-over (81%), while only 29 advertisements featured female voice-overs (19%), X 2 = 58.9, df = 1, p < 0.001). In British advertisements, 96 had male voice-overs (70%) and 42 had female voice-overs (30%), X 2 = 21.1, df = 1, p < 0.001). Comparison of the numbers of Saudi and British television advertisements that had voice-overs indicates that both samples were gender biased, which supports H6. Saudi advertisements, however, were the more gender biased in this respect, X 2 = 6.1, df = 1, p < 0.02.

Product Categories

An association emerged between gender and the types of products being advertised on Saudi television, but this was not in the case in British television advertisements. Small base sizes here meant that tests of statistical significance were conducted only among those categories with sufficient base sizes. On Saudi television advertisements, female lead figures were used significantly more often than male lead figures to promote personal care products. In all, 78% of personal care product advertisements were presented by women, compared with 22% by men, which supports H7. Further, the number of female leads (79%) in Saudi television advertisements for household cleaning products was significantly greater than those featuring male leads (21%). These two categories were aggregated to produce male and female cell sizes large enough for Chi-square analysis, X 2 = 10.13, df = 1, p < 0.01, which confirmed the significantly greater representation of female leads than male leads in advertisements for household and personal care products.

It is also clear that two further product categories featured greater numbers of male lead figures than female lead figures: beverage advertisements (Saudi advertisements: 85% and 15%; British advertisements: 60% and 40%) and car advertisements (Saudi advertisements: 100% and 0%; British advertisements: 62.5% and 37.5% respectively), which supports H8. In neither case, however, did cell sizes permit tests for statistical significance. Despite this, the raw data reveal clear gender differences. The zero return for female leads in car advertisements can be attributed to the fact that women in Saudi Arabia are not allowed to drive cars and therefore do not represent a target market for advertisers. There were no significant differences in use of male and female lead figures in relation to any product advertisement from British television, which offers no support for H7 or H8.

Finally, the prediction (H9) that gender stereotyping would be more pronounced in television advertisements in Saudi Arabia than in the UK was upheld. Hypotheses 2, 3, 4, 5, and 7 were all upheld for Saudi Arabian television advertisements, but not for those broadcast in the UK. H6 was proven in both countries, but was still more strongly upheld in the case if Saudi advertisements than British advertisements.

Discussion

Saudi society is male-oriented. Women have no voting privileges, are not allowed to drive cars, and are restricted in their manner of dress and their movements in public. This value system is grounded in Islamic teachings that assign men the role of guardians of their wives, children and unmarried sisters. Among many other duties, this guardianship includes providing protection and sufficient financial security. In consequence, the proportion of men in the workforce is greater than that of women (Delener 1994). In contrast, Britain is a more secular society in which gender equality has become the expectation in professional and private life (Assael 1998). There was an expectation therefore that the greater social and occupational role pluralism associated with women in the UK than Saudi Arabia might be reflected in the way each gender was represented in televised advertising (see Al-Makaty et al. 1996). This expectation was reinforced by the findings.

In this study, we compared the portrayal of men and women in television advertising in Saudi Arabia and Britain. On the basis of past research literature (e.g., Furnham and Mak 1999; Gunter 1995; Manstead and McCulloch 1981; McArthur and Resko 1975), it was anticipated that gender stereotyping would characterise television advertising, but because of their different value systems, that such stereotyping would be stronger in Saudi advertising than in British advertising.

The distribution of male and female lead characters in television advertising revealed no significant gender stereotyping on the basis of gender prevalence in lead roles in advertisements and very little difference between the Saudi Arabian and British commercial samples. These results for relative gender prevalence in television advertising are consistent with earlier American studies (Culley and Bennett 1976; Downs and Harrison 1985; McArthur and Resko 1975; O’Donnell W. J. and O’Donnell K. J. 1978), although our findings indicated a more even distribution in British television advertisements than earlier British studies (e.g., Furnham and Bitar 1993; Manstead and McCulloch 1981). In fact, the earlier British analyses had indicated that two out of three product representatives were male in the early 1980s (Manstead and McCulloch 1981), a figure that expanded to three out of four by the early 1990s (Furnham and Bitar 1993). The current study revealed a gender split that approached 50:50, which indicates an increase in the prevalence of women in lead roles in television advertising in Britain.

In the case of the Saudi sample, no comparable benchmark data exist with which to demonstrate any changes in relative prevalence of female and male leads in television advertising. Razzouk and Al-Khatib (1993) reported that 14% of their sample of Saudi television advertisements showed women in domestic roles. They used advertisements, rather than characters in advertisements as units of analysis, therefore no direct comparisons can be made between their data and the data from the current study.

In addition to visual appearances on screen, we also examined the gender distribution of voice-overs in television advertisements. In both Saudi and British advertisements, male voice-overs were significantly more prevalent than female voice-overs. The British findings confirmed the results of earlier British research (Furnham and Bitar 1993; Livingstone and Green 1986; Manstead and McCulloch 1981) and indeed similar research findings from many other countries (e.g., Furnham and Mak 1999; Neto and Pinto 1998). It is also important to note, however, that gender stereotyping on this dimension was significantly greater in Saudi Arabia than in Britain, which confirms the hypothesis that male bias would be stronger in Saudi advertising than in British advertising.

Previous research indicated that more pronounced gender stereotyping could be found in advertising in relation to the nature of appearances by men and women, as distinct from absolute numbers of appearances (e.g., Durkin 1985a, b, c; Geis et al. 1984). We found no gender differences in British television advertisements in relation to the roles occupied by characters on the screen, whereas in the Saudi sample, female leads were found significantly more often than male leads in domestic or household roles and less often in occupational or leisure activity roles.

These findings are only indirectly comparable with previous British research evidence because of differences in role definitions and measurement. Manstead and McCulloch (1981), for instance, distinguished between ‘autonomous roles’ (worker, professional, celebrity and interviewer/narrator) and ‘dependent roles’ (spouse, homemaker, boyfriend/girlfriend, and sex object). Furnham and Bitar (1993) also adopted this classification system. Results of both these studies indicate that the ‘dependent role’ was significantly more likely to be occupied by women than men. Furnham et al (2000) also indicated a clear predominance of women in ‘dependent roles’ in advertisements from another Muslim country—Indonesia, a finding that is consistent with our data from Saudi Arabia.

Another measure associated with role is setting. This measure also traces the extent to which lead characters—both men and women—are depicted in home-based or occupational or other locations. This factor has been regarded as an important indicator of gender bias in advertising (e.g., Bresnahan et al. 2001; Brettl and Cantor 1988; Furnham and Skae 1997). The findings here confirmed those for character roles by revealing significant gender distribution differences by location for television advertisements from Saudi Arabia, but not for those from Britain. In Saudi advertisements, women and girls were significantly more likely than men and boys to appear in home-based settings, whereas men and boys were significantly more likely than women and girls to appear in outdoor settings. There was greater gender stereotyping on this dimension in Saudi than in British advertisements, which lends further support to the anticipated differences between the two countries. The Saudi findings were consistent with those for another (non-Arab) developing Muslim nation (Indonesia; Furnham et al. 2000). The British findings, however, indicated less pronounced gender stereotyping than did earlier research in this country (e.g., Furnham and Bitar 1993; Manstead and McCulloch 1981), though differences in sampling time-frames may account in part for these discrepancies.

The age profiles of male and female lead characters differed in Saudi Arabian television advertisements; women exhibited a generally younger profile than did men. This finding was not replicated in British television advertisements. The Saudi, but not the British, results were consistent with earlier research (Furnham and Bitar 1993; Furnham et al. 2000; Schenider and Schneider 1979).

A link was found between gender and the types of products advertised for Saudi advertisements, though not for British advertisements. Women were strongly associated with personal care products and household cleaning products in Saudi advertisements, whereas men were most often associated with advertisements for beverages. The strong association of women with body and household products is consistent with previous research conducted in other Muslim environments (Malaysia—Bresnahan et al. 2001; Indonesia—Furnham et al. 2000). The non-significant links between gender and product type in the British advertisements also confirmed some results from earlier research for this country (Furnham and Bitar 1993). Earlier still, however, Manstead and McCulloch (1981) found a gender profile for product types (especially for body and household products) that resembled the current Saudi findings. Hence, there is evidence of a shift from significant gender stereotyping by product type in television advertising in Britain 20 years ago to a gender equitable profile that has been sustained over the last 10 years.

Our data indicate a continued presence of gender stereotyping in television advertising that is consistent with that recorded by other research for nearly 30 years. There is evidence of an overall difference between the two countries examined here, however. Saudi Arabian television advertising was found to display different patterns of male and female lead character portrayals from those in British television advertising, especially in relation to women’s representation in domestic roles rather than non-domestic roles, age, and types of products advertised.

The findings for the UK are consistent with the observations of other researchers that gender stereotyping appears to be on the decline in advertising in the Western world, but remains more firmly entrenched in advertisements in the developing world (Bresnahan et al. 2001; Furnham and Mak 1999; Furnham et al. 2000; Mwangi 1996).

These differences in gender representation may reflect important differences in cultural values between nations of which advertisers are aware, particularly when launching global or multi-national campaigns (Belk and Pollay 1985; Pollay 1983). Successful advertising campaigns in Saudi Arabia may be ones that adopt localised (see Shoham 1996) or ‘glocalised’ (see Whitelock and Chung 1989) strategies, whereby advertisements are tailored in whole or in part to ensure their acceptance by the country’s Islamic value system.