HIV
 

Social Science & Medicine Article in Press
 

Liberty Eaton, mailto:liberty_eaton@yahoo.co.uk mailto:liberty_eaton@yahoo.co.uk, a, Alan J. Flishera and Leif E. Aarøb

a Department of Psychiatry and Mental Health, University of Cape Town, Groote Schuur Hospital, Observatory 7925, South Africa
 

b Research Centre for Health Promotion, University of Bergen, Christiegt 12, N-5015, Norway

Available online 23 January 2002.
 

Article in Press - Uncorrected Proofs: these are articles that are peer reviewed and accepted articles to be published in this journal, but not yet finalized and that will be corrected by the authors. Therefore the text could change before final publication.
 

A growing body of evidence points to the complexity of sexual behaviour. HIV risk behaviour is influenced by factors at three levels: within the person, within the proximal context (interpersonal relationships and physical and organisational environment) and within the distal context (culture and structural factors). This paper presents the findings of a review of research on the factors promoting and perpetuating unsafe sexual behaviour in South African youth. Papers included in the review were dated between 1990 and 2000 and addressed sexual behaviour of youth between the ages of 14 and 35 years. Both published works and unpublished reports and dissertations/theses were included. The review concluded that at least 50% of young people are sexually active by the age of 16 years; the majority of school students who had ever experienced sexual intercourse reported at the most one partner in the previous year, with a persistent minority of between 1% and 5% of females and 10¯25% of males having more than four partners per year; and between 50% and 60% of sexually active youth report never using condoms. In terms of explanations for unsafe sexual behaviour among South African youth, the findings illustrate the powerful impact of the proximal and distal contexts, and in particular, the pervasive effect of poverty and social norms that perpetuate women's subordination within sexual relationships. Personal factors and the proximal and distal contexts interact to encourage HIV risk behaviour in ways that are not fully captured by social-cognitive models. The findings will be of interest to researchers and practitioners in the fields of adolescent sexual behaviour and HIV prevention in developing countries.

Author Keywords: Sexual behaviour; Adolescents; HIV/AIDS; South Africa
 

The poorest, most underdeveloped region in the world, Sub-Saharan Africa, faces by far the highest rate of HIV infections. Although this region accounts for only 10% of the world's population, 85% of AIDS deaths have occurred here (World Bank, 2000). Young people have the fastest-growing infection rates. In 1998, the HIV infection rates among South Africans aged 14¯19 years and 20¯24 years were 21.0% and 26.1%, with percentage increases from 1997 of 65.4% and 32.5%, respectively ( Adler & Qulo, 1999).

Several major theories of behaviour have been applied to understanding HIV-risk behaviour. These include the Health Belief Model (Becker and Becker; Janz & Becker, 1984; Rosenstock, 1966); the Theory of Reasoned Action ( Azjen & Fishbein, 1970) and its revised form, the Theory of Planned Behaviour ( Azjen, 1985); and Social Cognitive Learning Theory 1 (Bandura and Bandura). These theories (dubbed "social-cognitive" within the health psychology literature) mainly deal with factors within the triad: behaviour, personal factors, interpersonal factors and processes. (For an overview of such theories, see Conner and Norman, 1996.) One's behaviour is seen to be primarily a function of beliefs and subjective evaluations. The key cognitions and evaluations addressed by these theories include: vulnerability to a health risk; perceived severity of the health outcome; likelihood that changed behaviour will protect against the risk; confidence in changing one's behaviour effectively; the costs versus benefits associated with risky behaviour; perceived emotional and social consequences of heath-related behaviours; and perceptions about social norms (what other people think and feel, and whether the individual is motivated to comply with these perceived pressures). These variables may influence behaviour itself or the intention to behave in a certain manner ( Azjen & Fishbein, 1970).

These social-cognitive theories have been found to be valid and useful, especially within the contexts in which they were designed (that is, within Western societies). But they cannot be applied blindly in all circumstances and to all problems. This is particularly apparent in developing countries, where factors beyond the individual have an impact that warrants special consideration. Social-cognitive theories do recognise the relevance of factors beyond the individual. However, they tend to emphasise personal processes and the subjective aspects of social influences, to the neglect of the objective aspects of social influences and the distal societal and cultural context.

The need to consider objective social, economic, environmental and political factors has been recognised by AIDS researchers in Africa (Webb, 1997), as well as by the designers of health intervention models such as the PRECEDE¯PROCEDE model ( Green & Kreuter, 1991) and the PEN-3 model ( Airhihenbuwa, 1995). These models do not, however, offer predictive theories of behaviour, and do not suggest how social context interacts with factors at the individual and interpersonal levels of analysis.

If we wish to understand sexual risk behaviour in Southern Africa, we need to consider the interactive effects of factors at three levels: within the person, within his or her proximal context, and within the distal context. Personal factors include cognitions and feelings relating to sexual behaviour and HIV/AIDS, as well as thoughts about one's self (such as self-efficacy and self-esteem). The proximal context comprises interpersonal relationships and the physical and organisational environment. The distal context includes culture and structural factors. Culture comprises aspects such as traditions, the norms of the larger society, the social discourse within a society, shared beliefs and values, and variations in such factors across subgroups and segments of the population. Structural factors include legal, political, economic or organisational elements of society. The importance of cultural and structural factors and the neglect of such factors in health behaviour research have been recognised in recent publications (Cockerham, 1997; Dressler & Oths, 1997; Eakin, 1997). Fig. 1 presents a framework for organising the relationship between sexual behaviour, personal factors and the proximal and distal contexts.

As is the case with Social Cognitive Theory (Bandura, 1986) and a recent European version of value-expectancy models ( Kok, Schaalma, De Vries, Parcel, & Paulssen, 1996), the present model is meant to include both subjective and objective influences on behaviour. Since we shall apply the model to understanding influences on sexual behaviours specifically, we depict a one-way process where individuals and their immediate environment are influenced by broader social conditions. It should be noted, however, that in general terms all these factors are potentially reciprocally determining ( Bandura, 1977).

The question under consideration in the present review was this: Why is it that South African youth in the 1990s continued to practice unsafe sex (as evidenced in the spiralling rates of HIV infection), despite the concerted efforts of educational and HIV prevention campaigns to influence their behaviour? The aim was to integrate disparate research findings in order to derive a larger scale view of the factors that promote or perpetuate unsafe sexual behaviour in the South African context. We shall use the model described above to synthesise and discuss the data.

Studies dated between 1990 and 2000 were eligible for inclusion. We performed keyword searches for articles in English or Afrikaans on the following databases: Medline, Psychlit, Sociofile, Social Sciences Index, and the Index to South African Periodicals (ISAP). We scanned the reference lists of articles for further studies. We searched the South African Medical Journal by hand. We considered both published works and unpublished dissertations and reports. The main criterion for inclusion in the review was that the research subjects must have been between the ages of 14 and 35. Studies of adult populations were eligible if the study provided a breakdown of age groups within its sample, and people under 35 formed a substantial proportion of the sample.

We excluded studies that focused on particular adult occupational groupings because we felt the disproportionate number of studies reporting on health care and social service workers would bias the review findings. This decision meant that studies on other groups such as mineworkers, sex workers and truck drivers, in which some young people would incidentally have been included, were also excluded so as to retain the emphasis on the youth in general.

A total of 75 reports and papers were included in the review. The most relevant studies reporting quantitative data on sexual activity are summarised in Table 1.

The present review addresses the three types of sexual risk behaviour that have received the most research attention in South Africa: being sexually active (as opposed to abstaining from or postponing sexual activity); having many partners (either serially or concurrently); and practicing unprotected sex (which includes the irregular or incorrect use of condoms). Other risk or protective behaviours have received too little attention in the literature to warrant their discussion in the present review.

Table 1 lists the findings on the proportions of young people who are sexually experienced, the number of partners they have had, and their use of condoms. Despite the lack of accurate national figures on age of sexual debut, the studies suggest that at least 50% of young people in South Africa are sexually active by age 16, and probably 80% are by the age of 20. Boys report earlier sexual debut than do girls, and Black ("African") youth are more likely to start sexual activity in their teens than are other ethnic groups.

Research on the total number of partners had by sexually active adolescents shows large variability between samples, as well as using different types of questions to investigate this issue. Overall, the majority of school-going adolescents reported having one or two partners in their lifetime, and over 60% of university students reported no partner or one partner in the last year (Table 1). Thus most young people are not being promiscuous. But the studies found a persistent minority having more than four partners per year. Only 1¯5% of women fall within this group, compared to 10¯25% of men. The research also suggests that between 10% and 30% of sexually active young people have more than one sexual partner at a given time, with more men than women engaging in concurrent multiple partnering.

The majority of sexually active young people use condoms irregularly, if at all. In mixed-gender samples, a maximum of 86% of sexually active respondents report ever having used a condom, with the average being much lower. A maximum of 55%, and more likely under 20%, of young people use condoms at every sexual encounter. And an overall estimate of 50¯60% of youth in these studies (with a range of 23¯85%) report not using condoms at all (Table 1).

The statistics make it clear that young South Africans put themselves at risk for HIV infection through unprotected sex, starting in their teens. But the numbers do not explain the problem. Why is it that youth do not protect themselves better?

Many of the studies we reviewed reported reasons for unprotected sexual activity by youth. Because of their different methods (ranging from qualitative interviews to focus groups to questionnaires), it was not feasible to conduct a quantitative meta-analysis of the findings. Instead, the studies were read carefully and their key results or conclusions tabulated and compared. When synthesised in this way, the studies indicate a recurring pattern of conditions that leads to risky sexual behaviour. Although our method of review is open to the criticism of being potentially biased against papers representing views that dissent from our own, we believe that the papers do represent a largely unified picture.

The factors that promote risk behaviours or create barriers to safer practices will be structured according to three domains of analysis: personal factors; the proximal environment (including interpersonal factors, and the immediate living environment); and the broader social context (including structural and cultural factors). We will then illustrate how these three categories of factors interact with each other to influence sexual behaviour.

Over 90% of young South Africans in the 1990s knew that AIDS is a fatal, sexually transmitted disease (Everatt & Orkin, 1993; Richter, 1996; Strebel & Perkel, 1991; Van Wijk, 1994; Varga & Makubalo, 1996). However, understanding of the nature of HIV, the mechanisms of transmission and methods of prevention were not as good. Fewer than 50% of young people understood how HIV and AIDS are related ( Elkonin, 1993; Naidoo, 1994; Richter, 1996). In fairly basic pencil-and-paper tests of knowledge, young South Africans could answer roughly 50¯80% of the questions correctly (e.g. Blecher et al., 1995; Elkonin, 1993; Harvey, 1997; Peltzer, Cherian, & Cherian, 1998; Perkel, 1991; Strebel & Perkel, 1991; Van Aswegen, 1995; Visser & Moleko, 1999). This shows serious gaps in knowledge. In the few interview studies undertaken, it appeared from respondents' spontaneous answers that their understanding of HIV/AIDS and prevention options was sketchy (e.g. NPPHCN, 1996; Richter, 1996; Varga & Makubalo, 1996).

There was also uncertainty about the proper use of condoms. On closed-ended (multiple choice or true-false) questions, 70¯90% of respondents usually indicated that condoms can protect against AIDS (Blecher et al., 1995; Du Plessis et al., 1993; Elkonin, 1993; Friedland et al., 1991; Goliath, 1995; Govender et al., 1992; Naidoo, 1994; Strebel & Perkel, 1991; Van Dyk, 1994). In one study, open-ended questions elicited similar levels of knowledge about condoms ( Richter, 1996); but in another three studies, only 20¯40% of respondents could spontaneously cite condoms as a mode of protection ( NPPHCN, 1996; Ratsaka & Hirschowitz, 1995; Van Aswegen, 1995). And in a study where 88% of sexually active respondents agreed that condoms protect against AIDS, none actually used them ( Govender et al., 1992).

Serious misconceptions were held by some youth, for example that hormonal contraceptives and intrauterine contraceptive devices offer protection against HIV infection, or that the same condom may be used more than once (Blecher et al., 1995; Du Plessis et al., 1993; Govender et al., 1992; Van Aswegen, 1995). Mythical disadvantages were attached to condoms, such as the widespread belief that condoms can "disappear" into women and cause them serious injury (e.g. Abdool Karim et al., 1994; Harvey, 1997; Nicholas, 1998). Two studies showed that 7¯10% of the youth did not know what condoms are ( Blecher et al. (1995); NPPHCN (1996); see also Mathews et al. (1990)).

The Health Belief Model (HBM) and Social-Cognitive Learning Theory (SCLT) both stress the importance of perceptions about the seriousness of a health threat, perceptions about one's personal vulnerability to a health threat, and one's perceived ability to reduce one's risk, as key determinants of health behaviour. Low perceived personal vulnerability is a risk factor because it reduces the motivation to take the necessary precautions. The South African research does indicate that higher perceived vulnerability and anxiety about personal risk is linked to greater intended and actual sexual behaviour change (e.g. Strebel & Perkel, 1991; Van Aswegen, 1995; Van Wijk, 1994).

Unfortunately, many South African youth underestimate their risk for contracting HIV. Researchers have used various methods for assessing perceived risk, so findings are not easily aggregated. But it is clear that fewer than half of South African youth in the 1990s perceived any risk to themselves, and fewer than 20% perceived a high risk (Everatt & Orkin, 1993; Friedland et al., 1991; Govender et al., 1992; NPPHCN, 1996; Perkel, 1991; Ratsaka & Hirschowitz, 1995; Richter, 1996; Simpson, 1996; Strebel & Perkel, 1991; Van Wijk, 1994; Visser et al., 1995).

Perceptions of risk are unrealistically low in some groups with high rates of sexual activity and low condom use. For example, Blecher et al. (1995) found that fewer than 40% of their sample felt any risk to themselves from AIDS, and only 9% perceived a serious risk. The sample were outpatients at an STD clinic¯¯a group of people who had already contracted a sexually transmitted disease, and who reported a very low rate of condom use.

Some of the factors that reduce perceived vulnerability have been identified through qualitative, interview-based research. One such factor is the tendency to deny the presence of HIV/AIDS in one's own community, particularly in rural areas where there is still great stigma attached to AIDS. Denial of risk and of personal responsibility may be more prevalent among men than among women (MacPhail & Campbell, 2000), a gendered response that will be discussed later.

The expectation that one can successfully complete a behaviour, such as using a condom, is theorised to be an important predictor of whether one attempts the behaviour (Azjen, 1985; Bandura, 1991). Two South African studies with young adults suggest that self-efficacy for condom use is indeed linked to higher self-reported condom use ( Peltzer, 1999; Reddy et al., 2000), although the direction of causality is not proved by the correlational methods used.

The main social-cognitive models posit that people consider positive and negative features of preventive behaviours and the balance will influence their behaviour. For example, many young men claimed that abstinence or suppression of sexual desire leads to ill health (Meyer-Weitz, Reddy, Weijtz, Van den Borne, & Kok, 1998). A further perceived disadvantage of abstinence is that it prevents people from demonstrating their fertility by conceiving babies. The desire of traditionally minded young Black men in South Africa to prove their virility by fathering children, and the pressure felt by girls to prove their love and fertility by conceiving, has been well documented (e.g., Kau, 1991; Preston-Whyte & Zondi, 1991; Varga & Makubalo, 1996; Wood, Jewkes, & Maforah, 1997). It should be noted, however, that most urban youth with modern lifestyles and aspirations do not want to become parents too soon ( MacPhail & Campbell, 2000; Richter, 1996).

On the one hand, condoms carry the same disadvantage by being contraceptive. In South African studies, up to 45% of young men in certain samples claimed that "condoms waste sperm" (e.g. Reddy, Meyer-Weitz, van den Borne, & Kok, 1999). On the other hand, young people who want to avoid pregnancy perceive condoms to be less reliable than hormonal contraceptives. In long-term relationships where pregnancy is perceived to be more of a risk than HIV, condoms are used only occasionally as supplementary contraception (e.g. Wood & Foster, 1995). Thus condoms are rejected from two sides: they prevent conception for those wanting to prove virility/fertility, but are not perceived to be contraceptiveenough for those whose main concern is contraception.

The disadvantage of condoms cited most often is loss of pleasure. Several studies report young men (and some women) claiming that they like sex to be "skin on skin" (e.g., MacPhail & Campbell, 2000; NPPHCN, 1996; Reddy & Meyer-Weitz, 1997; Reddy et al., 1999; Richter, 1996). For young people who prize flesh-to-flesh sex, the immediate, salient costs of HIV-preventive behaviour are perceived to be high in relation to weakly felt personal risk.

Further disadvantages of condoms cited by the youth are that too many condoms are required for many rounds of sex; fear of condoms breaking or slipping; and awkwardness in purchasing condoms. In two studies these practical problems received greater endorsement than the physical and psychological problems of reduced pleasure and intimacy (Madu & Peltzer, 1999; Nicholas, 1998). For those who believe the myth that condoms may disappear into women, causing injury or death, this is obviously also a major disadvantage to condom use ( Reddy & Meyer-Weitz, 1997).

A noticeable absence in the literature is research on perceived benefits of abstinence, mutual monogamy, or condom use. Further research is needed to elucidate the importance and salience of both negative and positive values attached to preventive behaviour.

In the Theory of Planned Behaviour and the Theory of Reasoned Action, intentions are postulated to be the major determinants of health-related behaviour. Research in Western contexts bears this out to some extent. For example, intentions to use condoms have a moderate to strong correlation of 0.44 with self-reported condom use (Sheeran & Orbell, 1998), although of course such correlational research does not prove that intentions cause behaviour. The intention to abstain altogether from sex until marriage is expressed by a small minority, mostly women from conservative Christian backgrounds (e.g., Venter, 1995), and rarely by young men. The intention to be monogamous is expressed by many women (although by no means all). With young men, however, the picture is different: many consider monogamy to be just as undesirable as abstinence. This will be discussed in more detail later. Monogamy or reducing the number of partners is, however, considered more acceptable than using condoms. When people are asked how they intend to change their behaviour to reduce their risk of HIV infection, the intention to be faithful or to reduce the number of partners is reported far more often than the intention to use condoms ( Blecher et al., 1995; Strebel & Perkel, 1991).

South African research has found that low self-esteem is associated with earlier onset of sexual activity and having more sexual partners (Goliath, 1995; Perkel, Strebel, & Joubert, 1991). It has been hypothesised that a person with a poor sexual self-concept may rely on others for affirmation. This may lead him or her to search for external affirmation in multiple sexual encounters. South African research also indicates that young people with low self-esteem may be more concerned about what their partners think of them and with avoiding displeasure or rejection from partners than are people with more positive, self-affirming self-concepts ( Perkel, 1991). A person with low self-esteem is therefore more likely to think that condoms are offensive to their partner, to think that using condoms may make their partner think they are dirty, to be embarrassed about using condoms and to have a negative attitude towards condoms ( Perkel et al., 1991). Low self-esteem seems to undermine abstinence, monogamy and condom use.

Communication with one's partner about STD risk and condom use has been found to be strongly correlated with willingness to use condoms and self-reported use (Reddy et al., 2000). But talking about condoms is not easy. Discussions tend to be limited and awkward ( Reddy & Meyer-Weitz, 1997). Introducing condoms into a sexual encounter is perceived to break the intimacy and romance of the moment (e.g. Meyer-Weitz et al., 1998; Van Dyk, 1994; Wood & Foster, 1995). Both men and women perceive condoms to be associated with promiscuity, STDs and AIDS, so that suggesting condom use implies either that one has a sexually transmitted disease, or that one mistrusts one's partner ( Meyer-Weitz et al., 1998; NPPHCN, 1996; Richter, 1996; Van Dyk, 1994).

Within trusting relationships where contraception and sexual choices are discussed, condoms tend to be abandoned in favour of less intrusive, more effective contraceptives (Simpson, 1996; Wood & Foster, 1995). The act of leaving condoms behind symbolises a new level of commitment within the relationship. There is also an assumption that loving or long-term relationships necessarily involve less risk, even when the partner's HIV status is unknown ( Kelly & Parker, 2000). Condoms are reserved for casual encounters or ``secret lovers'' other than one's steady partner ( MacPhail & Campbell, 2000). It is extremely difficult to re-introduce condoms into a steady relationship once they have been abandoned, as this calls into question the assumption of mutual monogamy ( Wood & Foster, 1995).

But negotiating condom use can also be a positive experience. Open discussion and mutual agreement to behavioural change can strengthen the relationship, increase partners' respect for each other, confirm that they care about each other's well-being, and enhance their sexual intercourse by removing any anxiety about the risk of infection (Wood & Foster, 1995).

Sexual negotiation of any kind¯¯be it about condom use, faithfulness, or about the nature and frequency of sexual intercourse¯¯is lacking in many sexual relationships among young South Africans.

Qualitative research reports that young people's heterosexual relationships in certain communities frequently involve sexual coercion of, and violence towards, the female partner. Although numbers are not generally reported, these studies describe young women who are physically forced or bullied into having sex. Boyfriends, who claim that a romantic relationship must necessarily involve full penetrative sex when and how the man wants it, feel justified in using physical assault or threats of violence to coerce their girlfriends into having sex (Varga & Makubalo, 1996; Wood & Jewkes, 1997; Wood, Maforah, & Jewkes, 1998).

In such relationships the male partner largely controls sexual activity. The threat of violence or rejection prevents girls and women from insisting on condom use (Meyer-Weitz et al., 1998; Varga & Makubalo, 1996). While young women in such relationships may be violently punished for perceived unfaithfulness ( MacPhail & Campbell, 2000; Meyer-Weitz et al., 1998; Reddy & Meyer-Weitz, 1997; Whitefield, 1999), their boyfriends claim the right to have multiple sexual partners ( Meyer-Weitz et al., 1998; Reddy & Meyer-Weitz, 1997; Richter, 1996). In relationships with such an imbalance of power, young women's ability to practice safer sex is constrained by their partners' demands.

South African research has addressed the issue of peer pressure only in studies of Black youth. This research indicates that both girls and boys experience considerable same-sex peer pressure to be sexually active (Buga et al., 1996; Cassimjee, 1998; NPPHCN, 1996). For boys the pressure has to do with proving manliness, and having many sexual partners wins a young man status and admiration. Young men often encounter negative peer attitudes towards condoms ( Blecher et al., 1995; MacPhail & Campbell, 2000). For girls pressure sometimes comes from sexually experienced peers who exclude inexperienced girls from group discussions because they are still "children" ( Wood et al., 1997a).

Peer pressure does not, however, have the same negative influence on all youth. Individuals differ in their susceptibility (Perkel, 1991), and young men appear to be influenced to greater extent than are young women ( MacPhail & Campbell, 2000). Peer pressure is also not necessarily a negative influence. Positive examples set by friends and role models can promote safer sexual behaviour ( Perkel, 1991).

South African adolescents report poor communication with parents about sexual matters. Research participants have claimed that their parents refuse to talk to them, give them only vague injunctions rather than information, and may even punish them for raising the subject (Boult & Cunningham, 1991; Kau, 1991; Kelly, 2000; Kelly & Parker, 2000; Visser et al., 1995; Wood, Maepa, & Jewkes, 1997).

When there is poor communication within the family about sex, both supervision and lack of supervision from parents may contribute to unsafe sexual behaviour. When parents forbid contraception in an effort to control their children's sexual activity, the fear of discovery and parental anger leads to lower use of condoms (Wood et al., 1997b). Condoms are also often dispensed with when young people are hurriedly taking the opportunity to have sex while their parents are out ( MacPhail & Campbell, 2000).

But the opposite also happens. Particularly in rural areas, some mothers who feel they cannot control their children's sexual behaviour arrange for their daughters to receive contraceptive injections from menarche (Wood et al., 1997b). The resulting reduction in pregnancy risk, accompanied by low parental guidance and supervision, has contributed to increased rates of adolescent sexual activity and lower rates of condom use in these communities ( Kelly & Parker, 2000).

In addition to parents, staff at public health clinics could play a significant role in young people's sexual lives. These clinics are supposed to provide free condoms as well as general sexual health services. But young people report that clinic staff sometimes scold or mock them when they go for condoms (MacPhail & Campbell, 2000; Richter, 1996; Wood et al., 1997b). Clinic services are also perceived to lack privacy and confidentiality ( Abdool Karim, Preston-Whyte, & Abdool Karim, 1992b; Health Systems Development Unit, 1998; MacPhail & Campbell, 2000). Even friendly and efficient staff may show surprise when young people come to collect condoms, trying to persuade girls to use more reliable methods of contraception without consideration for protection against HIV, or being too embarrassed to offer advice about condom use ( Abdool Karim et al., 1992b). The result is that young people are deterred from making full use of the services available, and there are many lost opportunities for counselling and HIV/AIDS intervention.

Access does not appear to be a major problem as young people use a variety of sources to obtain condoms (MacPhail & Campbell, 2000). Access to free condoms may be restricted by the negative attitudes of clinic staff, discussed above, or by clinics running out of supplies ( Abdool Karim et al., 1992a). Lack of planning may be a greater problem, since young people do not always have a condom when they need one, and this does not deter them from having sex ( Disler, 1990; MacPhail & Campbell, 2000).

A recent study suggests that youth with high access to the media show a satisfactory level of knowledge about HIV/AIDS. Young people in rural and poor areas where media penetration is low clearly require alternative sources of information (Kelly, 2000).

Young people in areas with chronic unemployment and a lack of recreational facilities claim that they use sex as entertainment when there is nothing else for them to do (Kelly & Parker, 2000; Wood et al., 1997b).

For boys as well as girls on the street, sexual abuse and rape is a constant threat (Swart-Kruger & Richter, 1997). But voluntary sex is also a health risk. Prostitution is seen as being a good way to earn money as it pays more than other street activities, but clients usually want unprotected sex ( Swart-Kruger & Richter, 1997). Street youth also have romantic relationships involving sexual intercourse. Youth on the street often have more than one partner at a time and tend not to use condoms in those relationships ( Kruger & Richter, 1996). An additional risk factor is widespread substance abuse. Sexual intercourse while under the influence of glue, marijuana or alcohol reduces condom use ( Kruger & Richter, 1996).

Being in custody as a convicted or an awaiting-trial prisoner must be seen as a risk factor for HIV infection. De Ridder (1997) reports that the gang system operates just as strongly in the juvenile awaiting trial section of South African prisons as it does in the main prisons. Violence between members of gangs and victimisation of new inmates not affiliated to a gang are commonplace and can be vicious. Rape is a grim reality of youth in prison. Some boys report that the only way to overcome the humiliation of rape is to rape other boys ( De Ridder, 1997). Thus boys in prison are at risk for HIV infection from blood contact during a violent fight or assault, and from rape. We did not find any articles pertaining to voluntary intercourse involving youth in the prisons, but this may also be a risk factor.

As mentioned in the discussion of coercive sexual relationships, qualitative research points to the widespread low status of South African women within sexual relationships. The research has focused on Black communities, but this does not mean that women in other ethnic/cultural groups do not also experience oppression in their sexual relationships. (We did not come across any comparable research within other groups, or comparing ethnic groups.)

It has been observed elsewhere that traditional African cultures are frequently patriarchal and oppressive towards women (Airhihenbuwa, 1995). Writers on HIV/AIDS prevention in Africa have noted that pervasive, culturally entrenched gender discrimination increases the risk of HIV infection for African women (e.g. Ng'weshemi, Boerma, Bennett, & Schapink, 1997; Webb, 1997). But sexual coercion and violence within relationships is also linked to socio-economic status. Because ethnicity and socio-economic status are strongly linked in South Africa, we cannot assume that African cultural norms entirely explain the sexual subordination of Black South African women.

Research into the discourses that surround the subordination of women, reveals two main themes relating to male sexuality: biologically determined "need", and sexual "rights". For example, in discussions of monogamy, men in particular claim that they need variety (e.g. Meyer-Weitz et al., 1998; NPPHCN, 1996; Richter, 1996). They claim that it is in man's nature to want many partners, and that staying with one woman therefore goes against the essence of being a man. Some women come to believe this, too. As one female respondent put it, "I know that a man cannot stop going to the street [to find other women]" ( Meyer-Weitz et al., 1998, p. 44). The notion that masculinity implies having unprotected sex with numerous partners is "particularly well-developed in South Africa" ( MacPhail & Campbell, 2000).

Likewise, young people justify impulsive, unprotected sex through a discourse of biology and desire. As mentioned earlier, many young men claim that they need sex in order to stay healthy. There is also a pervasive belief that sexual desire is a natural force that one should not attempt to control (Meyer-Weitz et al., 1998). This reasoning is applied mostly to male sexuality ( Cassimjee, 1998), although it is also used to justify sexual impulsiveness in women. This kind of thinking may be part of the problem of young people failing to postpone sexual encounters when condoms are not available ( Disler, 1990).

The discourse of rights appears in the way young men claim ``ownership'' of their sexual partners. It has been discussed how many young men feel justified in forcing their girlfriends into having sex. This behaviour is supported by a social norm that a man has a right to sexual intercourse within a romantic relationship, and that he therefore has the right to use force if necessary to obtain it. An illustrative finding comes from the extensive NPPHCN (1996) study, which found that only about one-third of adolescent boys in the survey believed their girlfriends had the right to refuse sex.

This set of norms about sexual domination is further entrenched by the belief that violent, sexually coercive behaviour is a sign of passion. A study of high school pupils found that 52% of the boys and 45% of the girls attributed rape or sexual harassment to love of one's partner (Whitefield, 1999). This conception of love serves to justify and perpetuate unacceptable levels of abuse within young people's relationships.

It is important to note, however, that not all young people accept these norms. Religious youth are more likely to postpone sexual activity and have fewer partners (Nicholas & Durrheim, 1995). There is a minority of young men who openly challenge the idea of coercive sex, and who insist on a woman's right to equal say in a relationship ( MacPhail & Campbell, 2000), likewise some young women express respect and admiration for those women who fight back and resist coercion. There is also some evidence for norms limiting male freedom of choice. In the study by Reddy et al. (1999) over 40% of both male and female respondents did not accord either women or men the right to refuse sex. These findings illustrate the point that social discourses of ``needs'' and ``rights'' are complex and contested, rather than monolithic within society.

South African research has rarely attempted a systematic analysis of urban¯rural differences with regards to sexual risk behaviour. Studies are mostly either based in an urban (e.g. Flisher et al., 1993; Flisher & Chalton, 1995; Nicholas, 1998; Peltzer, 1999; Reddy et al., 2000; Richter, 1996; Van Wijk, 1994) or a rural (e.g. Buga et al., 1996; Van Aswegen, 1995) setting. The few studies that have covered the whole country (e.g. Du Plessis et al., 1993; NPPHCN, 1996; Van Dyk, 1994) do not look for distinctions between urban and rural sexual behaviour patterns, or factors pertinent to sexual behaviour. Only three studies attempted an urban¯rural comparison ( Abdool-Karim et al., 1994; Goliath, 1995; Kelly, 2000). Taken together, they provide evidence of urban youth being better informed about HIV/AIDS than their rural peers. Poor, rural communities have less access to the media, and therefore need more information and are more dependent on outside experts for their information ( Kelly, 2000). Rural areas feature less HIV-preventive behaviour than semi-urban or urban areas ( Kelly, 2000). With regards to partner turnover and multiple concurrent partners, studies show no clear pattern of differences between urban and rural sites ( Abdool-Karim et al., 1994; Kelly & Parker, 2000). There is a need for research that identifies the most salient dimensions of difference between urban and rural communities, with regards to HIV risk.

It has been suggested that in South Africa "the most significant vector for predicting sustained adoption of risk prevention measures is socio-economic background" (Kelly & Parker, 2000, p. 27). In South African research, poverty, unemployment, overcrowding, and low levels of education certainly appear to be linked to higher levels of adolescent sexual activity and less knowledge about HIV and AIDS (e.g. Du Plessis et al., 1993; Preston-Whyte & Zondi, 1991; Wood et al., 1997b). Given the racialised social stratification that still characterises South Africa, problems associated with poverty mostly affect Black youth.

Poverty is often the reason for the commodification of sex, in which women in dire economic circumstances agree to sexual relationships with men in exchange for financial support (Adams & Marshall, 1998). In other instances the situation may not be so overtly an exchange, but an older boyfriend with money offers both status and the kinds of gifts and financial assistance that parents cannot afford ( Kelly & Parker, 2000). The exchange of sex for money or gifts means that sex happens on the man's terms¯¯which usually means without a condom ( Abdool Karim, 1998; Adams & Marshall, 1998). From the woman's perspective, protection from possible future illness may be a lower priority than meeting immediate economic needs. The research does not give a clear indication of the prevalence of commercial or survival sex among young people. For example, one study suggested both that commercialised sex is the exception to the rule, and that money is one of the dominant reasons for young women to have sexual relationships (MacPhail & Campbell, 2000).

Socio-economic status is also related to the likelihood of young people experiencing physical abuse and sexual coercion within relationships. One study of high school pupils found that adolescents with lower socio-economic status (SES) experienced eight times as much physical abuse and four times as much attempted rape and actual rape within relationships than did adolescents with high SES (Whitefield, 1999). Thus the sexual domination of young women by their partners, discussed under the section on interpersonal relationships, appears to happen more often in poor communities.

The discussion thus far has identified factors that influence young South Africans' sexual behaviour, and has arranged these factors within the three domains of our model. The personal factors and proximal and distal contexts have been kept separate in order to illustrate how each contributes to our understanding of sexual risk behaviour. But of course these three domains overlap to a certain extent and reciprocally influence each other. A few illustrations will suffice to show how these domains are related.

An attitude of bravado and denial of personal vulnerability to HIV can be seen as a personal risk factor. But research suggests that this response to HIV is much more common in young men than in women (MacPhail & Campbell, 2000), perhaps because of social norms that encourage young men to show sexual bravado, defiance of risk, and high levels of sexual activity ( Kelly & Parker, 2000; Reddy & Meyer-Weitz, 1997). Yet, girls may be persuaded by their boyfriends or even by girlfriends that the advantages of spontaneous unprotected sex outweigh the disadvantages. Without alternative adult supports or salient community norms, girls may be reluctant to lose a boyfriend by questioning their evaluation of unprotected sex.

<br>

Likewise, the difficulties that people experience in negotiating condom use at the interpersonal level are made worse by personal factors, such as self-esteem and self-efficacy for condom use, and cultural taboos against frank sexual discussion between men and women (Meyer-Weitz et al., 1998).

Lastly, it is worth reiterating the important connection between poverty and other risk factors. With regard to the proximal context, poverty is undoubtedly implicated in the homeless street life of some South African youth, and also arguably increases the chance of being held in prison¯¯two proximal contexts that hold elevated risk for HIV infection. Poor communities have fewer recreation facilities for their young people and less access to the media, thereby increasing sexual activity and reducing access to information about HIV (Kelly & Parker, 2000). With respect to other aspects of the distal context, poverty may also be linked to the discourses that support an unequal distribution of sexual power between men and women. For example, one study reported that adolescent girls from advantaged socio-economic backgrounds rejected sexist beliefs and discourses about relationships, while girls with lower socio-economic status supported these beliefs to the same extent as their male peers ( Whitefield, 1999). In the South African context, poverty as a structural factor exerts a pervasive influence on young people's sexual behaviour and HIV risk.

MacPhail and Campbell (1999); Wenger (1999)

References

Abdool Karim (1998)Q. Abdool Karim , Women and AIDS: The imperative for a gendered prognosis and prevention policy. Agenda 39 (1998), pp. 15¯25.

Abdool Karim et al. (1992a)S.S. Abdool Karim, Q. Abdool Karim, E. Preston-Whyte and N. Sankar , Reasons for lack of condom use among high school students. South African Medical Journal 82 (1992), pp. 107¯110. Abstract-MEDLINE

Abdool Karim et al. (1994)Q. Abdool Karim, N. Morar, N. Zuma, Z. Stein and E. Preston-Whyte , Women and AIDS in Natal/KwaZulu: Determinants of the adoption of HIV-protective behaviour. Urbanisation and Health Newsletter 30 (1994), pp. 3¯9.

Abdool Karim, Preston-Whyte, & Abdool Karim (1992b)Q. Abdool Karim, E. Preston-Whyte and S.S. Abdool Karim , Teenagers seeking condoms at family planning services, Part 1: A user's perspective. South African Medical Journal 82 (1992), pp. 107¯110.

Adams & Marshall (1998)H. Adams and A. Marshall , Off target messages¯¯poverty, risk and sexual rights. Agenda 39 (1998), pp. 87¯92.

Adler & Qulo (1999)G. Adler and O. Qulo , HIV/AIDS and STD's. In: N. Crisp and A. Ntuli, Editors, South African health review, Health Systems Trust, Durban (1999), pp. 301¯313.

<br>

Akande (1997)A. Akande , Black South African adolescents' attitudes towards AIDS precautions. School Psychology International 18 (1997), pp. 325¯341. Abstract-PsycINFO

<br>

Airhihenbuwa (1995)C.O. Airhihenbuwa Health and culture: Beyond the western paradigm, Sage, Thousand Oaks (1995).

<br>

Azjen (1985)I. Azjen , From intention to actions: A theory of planned behavior. In: J. Kuhl and J. Beckman, Editors, Action-control: From cognition to behavior, Springer, Berlin (1985), pp. 11¯39.

<br>

Azjen & Fishbein (1970)I. Azjen and M. Fishbein , The prediction of behavior from attitudinal and normative beliefs. Journal of Personality and Social Psychology 6 (1970), pp. 466¯487.

<br>

Bandura (1977)A. Bandura Social learning theory, Prentice-Hall, Englewood Cliffs, NJ (1977).

<br>

Bandura (1986)A. Bandura Social foundations of thought and action: A social cognitive theory, Prentice-Hall, Englewood Cliffs, NJ (1986).

<br>

Bandura (1991)A. Bandura , A social cognitive approach to the exercise of control over AIDS infection. In: R. DiClemente, Editor, Adolescents and AIDS: A generation in jeopardy, Sage, Beverly Hills (1991), pp. 1¯20.

<br>

Becker (1974)M.H. Becker , The health belief model and personal health behavior. Health Education Monographs 2 (1974), pp. 324¯508.

<br>

Becker (1988)M.H. Becker , AIDS and behavior change. Public Health Reviews 16 (1988), pp. 1¯11. Abstract-MEDLINE | Abstract-EMBASE

<br>

Blecher et al. (1995)M.S. Blecher, M. Steinberg, W. Pick, M. Hennick and N. Durcan , AIDS¯¯knowledge, attitudes and practices among STD clinic attenders in the Cape Peninsula. South African Medical Journal 18 (1995), pp. 1281¯1286. Abstract-MEDLINE | Abstract-EMBASE | Abstract

<br>

1] (1994). Bodibe, C. R. (1994). Investigating the sexual knowledge, attitudes and behaviour of black adolescents. Unpublished doctoral thesis, University of South Africa.

Boult & Cunningham (1991)B.E. Boult and P.W. Cunningham , Black teenage pregnancy in Port Elizabeth. Early Child Development and Care 75 (1991), pp. 1¯70. Abstract-PsycINFO

Buga et al. (1996)G.A.B. Buga, D.H.A. Amoko and D.J. Ncayiyana , Adolescent sexual behaviour, knowledge and attitudes to sexuality among school girls in Transkei, South Africa. East African Medical Journal 73 (1996), pp. 95¯100. Abstract-MEDLINE | Abstract-GEOBASE

Cassimjee (1998)Cassimjee, N. (1998). The role of gender in black adolescent sexuality: An ecosystemic approach. Unpublished masters dissertation, University of Pretoria.

Cockerham (1997)W.C. Cockerham , Lifestyles, social class, demographic characteristics, and health behavior. In: D.S. Gochman, Editor, Handbook of health behavior research, Vol. I, Personal and social determinants, Plenum Press, New York/London (1997), pp. 253¯265. Abstract-PsycINFO

Conner & Norman (1996)M. Conner and P. Norman Predicting health behaviour, Open University Press, Buckingham, UK (1996).

De Ridder (1997)T. De Ridder , Boys pay unbearable price to survive prison. In: S.R. Boikanyo and P. Donnell, Editors, Children and youth at risk into the 21st century, Masakhane Youth Consultancy, Pietermaritzburg (1997).

Disler (1990)Disler, S. A. (1990). Health related lifestyles of adolescents. Unpublished masters dissertation, University of Cape Town.

Dressler & Oths (1997)W.W. Dressler and K.S. Oths , Cultural determinants of health behavior. In: D.S. Gochman, Editor, Handbook of health behavior research, Vol. I, Personal and social determinants, Plenum Press, New York/London (1997), pp. 359¯378. Abstract-PsycINFO

Du Plessis et al. (1993)G.E. Du Plessis, A.J. Meyer-Weitz and M. Steyn Study of knowledge, attitudes, perceptions and beliefs regarding HIV and AIDS (KABP) among the general public, Human Sciences Research Council, Pretoria (1993).

Eakin (1997)J.M. Eakin , Work-related determinants of health behaviour. In: D.S. Gochman, Editor, Handbook of health behavior research. Vol. I. Personal and social determinants, Plenum Press, New York/London (1997), pp. 337¯357. Abstract-PsycINFO

Elkonin (1993)Elkonin, D. S. (1993). Acquired immune deficiency syndrome: Knowledge, attitudes, and sexual activity among university students. Unpublished masters dissertation, University of Port Elizabeth.

Everatt & Orkin (1993)Everatt, D., & Orkin, M. (1993). "Growing up tough": A national survey of South African youth. CASE National Youth Survey for the Joint Enrichment Project (JEP). Johannesburg: Community Agency for Social Enquiry.

Flisher & Chalton (1995)A.J. Flisher and D.O. Chalton , High-school dropouts in a working-class South African community: Selected characteristics and risk-taking behaviour. Journal of Adolescence 18 (1995), pp. 105¯121. Abstract-PsycINFO | Abstract-EMBASE | Abstract | Full-text via CrossRef

1] (1992). A.J. Flisher, M.M. Roberts and R.J. Blignaut , Youth attending Cape Peninsula day hospitals: Sexual behaviour and missed opportunities for contraceptive counselling. South African Medical Journal 82 (1992), pp. 104¯106. Abstract-MEDLINE | Abstract-EMBASE

Flisher et al. (1993)A.J. Flisher, C.F. Ziervogel, D.O. Chalton, P.H. Leger and B.A. Robertson , Risk-taking behaviour of Cape Peninsula high-school students, Part VIII: Sexual behaviour. South African Medical Journal 83 (1993), pp. 495¯497. Abstract-MEDLINE | Abstract | Abstract-EMBASE

Friedland et al. (1991)R.H. Friedland, S.K. Jankelowitz, M. De Beer, C. De Klerk, V. Khoury, T. Csizmadia, G.N. Padayachee and S. Levy , Perceptions and knowledge about the acquired immunodeficiency syndrome among students in university residences. South African Medical Journal 79 (1991), pp. 149¯154. Abstract-MEDLINE | Abstract-EMBASE

Goliath (1995)Goliath, C. G. (1995). Sekondereskoolleerlinge se persepsies van VIGS en VIGS-opvoeding. [Secondary school pupils' perceptions of AIDS and AIDS education]. Unpublished doctoral thesis, University of Port Elizabeth (in Afrikaans).

Govender et al. (1992)V. Govender, R. Bhana, A. Pillay, R. Panchia, G.N. Padayachee and M. De Beer , Perceptions and knowledge about AIDS among family planning clinic attenders in Johannesburg. South African Medical Journal 81 (1992), pp. 71¯73.

Green & Kreuter (1991)L.W. Green and M.W. Kreuter Health promotional planning: An educational and environmental approach ((2nd ed). ed.),, Mayfield, Palo Alto (1991).

Harvey (1997)Harvey, B. M. (1997). A quantitative survey of knowledge, attitudes and behaviour related to AIDS/HIV among Zulu-speaking standard eight high school students. Unpublished masters dissertation, Rhodes University, Grahamstown.

Health Systems Development Unit (1998)Health Systems Development Unit (1998). Adolescent sexuality and reproductive health in the Northern Province. Johannesburg: Department of Community Health, University of the Witwatersrand.

Janz & Becker (1984)N.K. Janz and M.H. Becker , The health belief model: A decade later. Health Education Quarterly 11 (1984), pp. 1¯47. Abstract-MEDLINE | Abstract-PsycINFO

Kau (1991)M. Kau , Sexual behaviour and knowledge of adolescent males in the Molopo region of Bophuthatswana. Curationis 14 (1991), pp. 37¯40. Abstract-MEDLINE

Kelly (2000)Kelly, K. (2000). Communicating for action: A contextual evaluation of youth responses to HIV/AIDS. Pretoria: Department of Health (Sentinel Site Monitoring and Evaluation Project for the Beyond Awareness Campaign).

Kelly & Parker (2000)Kelly, K., & Parker, P. (2000). Communities of practice: Contextual mediators of youth response to HIV/AIDS. Pretoria: Department of Health (Sentinel Site Monitoring and Evaluation Project for the Beyond Awareness Campaign).

Kok, Schaalma, De Vries, Parcel, & Paulssen (1996)G. Kok, H. Schaalma, H. De Vries, G. Parcel and T. Paulssen , Social psychology and health education. European Review of Social Psychology 7 (1996), pp. 241¯282.

Kruger & Richter (1996)J.M. Kruger and L.M. Richter , South African street children: At risk for AIDS?. Africa Insight 26 (1996), pp. 237¯243.

MacPhail and Campbell (1999)C. MacPhail and C. Campbell , Evaluating HIV/STD interventions in developing countries: Do current indicators do justice to advances in prevention techniques?. South African Journal of Psychology 29 (1999), pp. 149¯165.

MacPhail & Campbell (2000)C. MacPhail and C. Campbell , `I think condoms are good but, AAI, I hate those things': Condom use among adolescents and young people in a Southern African township. Social Science and Medicine 52 (2000), pp. 1613¯1627.

Madu & Peltzer (1999)S.N. Madu and K. Peltzer , Complaints of university students about male use of condoms as a preventive measure against HIV/AIDS in South Africa. Journal of Psychology in Africa 9 (1999), pp. 78¯90. Abstract-PsycINFO

Makiwane (1998)M.B. Makiwane , Adolescent pregnancy and reproductive health in Transkei (rural South Africa). African Journal of Reproductive Health 2 (1998), pp. 41¯48. Abstract-MEDLINE

Mathews et al. (1990)C. Mathews, L. Kuhn, C.A. Metcalf, G. Joubert and N.A. Cameron , Knowledge, attitudes and beliefs about AIDS in township school students in Cape Town. South African Medical Journal 78 (1990), pp. 511¯516. Abstract-MEDLINE | Abstract-EMBASE

Meyer & Le Roux (1994)J.C. Meyer and J.A. Le Roux , The attitude towards nature, and incidence of sexual intimacy among students at a South African university. Journal of Sex Education and Therapy 20 (1994), pp. 255¯263. Abstract-PsycINFO

Meyer-Weitz, Reddy, Weijtz, Van den Borne, & Kok (1998)A. Meyer-Weitz, P. Reddy, W. Weijtz, B. van den Borne and G. Kok , The socio-cultural contexts of sexually transmitted diseases in South Africa: Implications for health education programmes. AIDS Care 10 Suppl. 1 (1998), pp. s39¯s55.

Naidoo (1994)Naidoo, S. (1994). Acquired immune deficiency syndrome: Knowledge, attitudes, and sexual activity among black adolescents. Unpublished masters dissertation, University of Port Elizabeth.

Ng'weshemi, Boerma, Bennett, & Schapink (1997)Ng'weshemi, J., Boerma, T., Bennett, J., & Schapink, D. (Eds.), (1997). HIV prevention and AIDS care in Africa: A district level approach. Amsterdam: Royal Tropical Institute.

Nicholas (1998)L.J. Nicholas , Black South African students' beliefs and attitudes about condoms. Psychological Reports 83 (1998), pp. 891¯894. Abstract-MEDLINE | Abstract-PsycINFO

Nicholas & Durrheim (1995)L. Nicholas and K. Durrheim , Religiosity, AIDS, and sexuality knowledge, attitudes, beliefs, and practices of black South African first-year university students. Psychological Reports 77 (1995), pp. 1328¯1330. Abstract-MEDLINE | Abstract-PsycINFO

NPPHCN (1996)NPPHCN (1996). Youth speak out...for a healthy future: A study of youth sexuality. Johannesburg: National Progressive Primary Health Care Network (NPPHCN).

Peltzer (1999)K. Peltzer , Factors affecting condom use in an urban adult community of the Northern Province, South Africa. Journal of Psychology in Africa 9 (1999), pp. 66¯77. Abstract-PsycINFO

Peltzer, Cherian, & Cherian (1998)K. Peltzer, L. Cherian and V.J. Cherian , AIDS awareness of secondary school pupils in the Northern Province of South Africa. Psychological Reports 83 (1998), pp. 955¯958. Abstract-MEDLINE | Abstract-PsycINFO

Perkel (1991)Perkel, A. K. (1991). Psychosocial variables in the transmission of AIDS. Unpublished doctoral thesis, University of the Western Cape.

Perkel, Strebel, & Joubert (1991)A.K. Perkel, A. Strebel and G. Joubert , The psychology of AIDS transmission: Issues for intervention. South African Journal of Psychology 21 (1991), pp. 148¯152. Abstract-PsycINFO

Preston-Whyte & Zondi (1991)E. Preston-Whyte and M. Zondi , Adolescent sexuality and its implications for teenage pregnancy and AIDS. Continuing Medical Education 9 (1991), pp. 1389¯1394.

Ratsaka & Hirschowitz (1995)M. Ratsaka and R. Hirschowitz , Knowledge, attitude and beliefs amongst inhabitants of high density informal settlements with regard to sexuality and AIDS in Alexandra township. Curationis 18 (1995), pp. 41¯44. Abstract-MEDLINE

Reddy & Meyer-Weitz (1997)Reddy, P., & Meyer-Weitz, A. (1997). Sacred or secret¯¯psychosocial and contextual determinants of STD-related behaviour. South African Research Council (Technical report).

Reddy, Meyer-Weitz, van den Borne, & Kok (1999)P. Reddy, A. Meyer-Weitz, B. van den Borne and G. Kok , STD-related knowledge, beliefs and attitudes of Xhosa-speaking patients attending STD primary health-care clinics in South Africa. International Journal of STD & AIDS 10 (1999), pp. 392¯400. Abstract-MEDLINE | Abstract-PsycINFO | Abstract-EMBASE | Abstract-Elsevier BIOBASE

Reddy et al. (2000)P. Reddy, A. Meyer-Weitz, B. Van den Borne and G. Kok , Determinants of condom-use behaviour among STD clinic attenders in South Africa. International Journal of STD & AIDS 11 (2000), pp. 521¯530. Abstract-MEDLINE | Abstract-PsycINFO | Abstract-EMBASE | Abstract-Elsevier BIOBASE

Richter (1996)Richter, L. M. (1996). A survey of reproductive health issues among urban black youth in South Africa (Final grant report). Centre for epidemiological Research in South Africa, Pretoria.

Rosenstock (1966)I.M. Rosenstock , Why people use health services. Millbank Memorial Fund Quarterly 44 (1966), pp. 94¯124.

Sheeran & Orbell (1998)P. Sheeran and S. Orbell , Do intentions predict condom use? Meta-analysis and examination of six moderator variables. British Journal of Social Psychology 37 (1998), pp. 231¯250. Abstract-MEDLINE | Abstract-PsycINFO

Simpson (1996)Simpson, M. R. (1996). Knowledge of safe sex practices and HIV transmission, propensity to risk taking, and alcohol/drug use in the aetiology of unprotected sex. Unpublished masters dissertation, Rhodes University, Grahamstown.

<br>

Smith et al. (1998)A.M.A. Smith, R. De Visser, A. Akande, D. Rosenthal and S. Moore , Australian and South African undergraduates' HIV-related knowledge, attitudes, and behaviors. Archives of Sexual Behavior 27 (1998), pp. 279¯294. Abstract-MEDLINE | Abstract-PsycINFO | Abstract-EMBASE | Abstract

<br>

Strebel & Perkel (1991)Strebel, A., & Perkel, A. (1991). "Not our problem": AIDS knowledge, attitudes, practices and psychosocial factors at UWC. Psychology Resource Centre Occasional Paper Series, No. 4, University of the Western Cape.

<br>

Swart-Kruger & Richter (1997)J. Swart-Kruger and L.M. Richter , AIDS-related knowledge, attitudes and behaviour among South African street youth: Reflections on power, sexuality and the autonomous self. Social Science and Medicine 45 (1997), pp. 957¯966. Abstract | Journal Format-PDF (724 K)

<br>

Van Aswegen (1995)Van Aswegen, E. (1995). Aids-related knowledge: Attitudes and behavioural practices among high-school pupils. S.A. Family Practice, May 1995, 307¯318.

<br>

Van Dyk (1994)A. Van Dyk , Vigs-voorligting: Kennis, houdings en gedrag in `n multi-kulturele, heteroseksuele konteks. [AIDS counselling: knowledge, attitudes and behaviour in a multi-cultural, heterosexual context]. Koers 59 (1994), pp. 221¯245 (in Afrikaans) .

Van Wijk (1994)Van Wijk, B. (1994). The dynamics of HIV transmission amongst a group of school-going adolescents in South Africa. Unpublished honours project, University of Cape Town.

Varga & Makubalo (1996)C. Varga and L. Makubalo , Sexual non-negotiation. Agenda 28 (1996), pp. 31¯38.

Venter (1995)A. Venter , Voorhuwelikse seksuele permissiwiteit onder damestudente. [Premarital sexual permissiveness among female students.]. South African Journal of Sociology 26 (1995), pp. 143¯149 (in Afrikaans) .

Visser & Moleko (1999)M. Visser and A. Moleko , High-risk behaviour of primary school learners. Urban Health and Development Bulletin 2 (1999), pp. 69¯77.

Visser et al. (1995)M.J. Visser, J.L. Roos and L. Korf , AIDS prevention on the campus. South African Journal of Higher Education 9 (1995), pp. 165¯173.

Webb (1997)D. Webb HIV and AIDS in Africa, David Philip, Cape Town (1997).

Wenger (1999)E. Wenger Communities of practice: Learning, meaning and identity, Cambridge University Press, Cambridge (1999).

Whitefield (1999)Whitefield, V. J. (1999). A descriptive study of abusive dating relationships amongst adolescents. Unpublished masters dissertation, University of Cape Town.

Wood & Foster (1995)C. Wood and D. Foster , "Being the type of lover...": Gender-differentiated reasons for non-use of condoms by sexually active heterosexual students. Psychology in Society 20 (1995), pp. 13¯35.

Wood & Jewkes (1997)K. Wood and R. Jewkes , Violence, rape, and sexual coercion: Everyday love in a South African township. Gender and Development 5 (1997), pp. 41¯46. Abstract-GEOBASE

Wood, Jewkes, & Maforah (1997a)K. Wood, R. Jewkes and F. Maforah , The violence connection in reproductive health: Teenage accounts of sexual relationships in Khayelitsha. Urbanisation and Health Newsletter 34 (1997), pp. 21¯24.

Wood, Maepa, & Jewkes (1997b)Wood, K., Maepa, J., & Jewkes, R. (1997b). Adolescent sex and contraceptive experiences: Perspectives of teenagers and clinic nurses in the Northern Province. Pretoria: Centre for Epidemiological Research in South Africa (Women's Health).

Wood, Maforah, & Jewkes (1998)K. Wood, F. Maforah and R. Jewkes , "He forced me to love him": Putting violence on adolescent sexual health agendas. Social Science and Medicine 47 (1998), pp. 233¯242. SummaryPlus | Article | Journal Format-PDF (184 K)

World Bank (2000)World Bank (2000). Intensifying action against HIV/AIDS: Responding to the development crisis. Washington, DC: World Bank.

Corresponding author. Current address: School of Psychology, University of St Andrews, Fife, Scotland KY16 9JU, UK; email: liberty_eaton@yahoo.co.uk

1 "Social cognitive theories" is used to denote a whole range of theories and models, while "Social Cognitive Theory" since 1986 refers to Bandura's theory of behaviour.

© Speak Out Terms of use