Sexual perceptions and practices of young people in Northern Thailand

Abstract

This study draws together survey and qualitative data on sexual practices among more than 1,750 young Northern Thai people aged 17-20 years. The survey data indicate that sexually active young people frequently engage in, or are subjected to, risk-taking behaviours that may expose them to sexually transmitted infections and unwanted pregnancies. These include having multiple sexual partners and quite frequent partner turnover. High percentages also engage in unprotected sexual intercourse with various types of sexual partner (steady, casual and paid), and young women especially had often experienced sexual coercion. Qualitative data revealed a mixture of perceptions and practices affecting sexual intercourse among the young, such as having unplanned sex, engaging in sexual relations to display love or cement committed relationships, and having serial relationships, both monogamous and non-monogamous. We conclude that condom use should be a central focus of activities aimed at preventing adverse sexual health outcomes, but that new intervention approaches to encourage use of other contraceptives are also needed. Changes in sexual norms among young people also need to be acknowledged and accepted by older Thai generations in order for programs and interventions to combat negative sexual and reproductive health consequences to be more effective.

Introduction

For some decades now, substantial proportions of adolescents in most Western countries have engaged in premarital sex. A study in the United States of trends in sexual activity among young American women from 1982 to 1995 (Singh and Darroch 1999) revealed that throughout that period about 40 percent of 15-19 year old women (and 27-28 percent of 15-17 year-olds) had had sexual intercourse in the last 3 months. Santelli et al. (2000) likewise reported 1995 estimates from three national surveys of percentages of 15-17 year old US women who had ever had intercourse ranging from 36.5 percent to 52.1 percent, with male estimates ranging from 41.3 percent to 52.9 percent. In a study of 24 developed countries (not including the US) conducted in 2002 (Godeau et al. 2008), levels of intercourse experience among national samples of 15 year old students ranged from 17.2 percent (Spain) to 47.1 percent (Ukraine) for males (median 24.6 percent) and from 2.7 percent (Macedonia) to 39.9 percent (England) for females (median 18.5 percent). Another comparative study of 1991-96 data for Sweden, France, Canada, Great Britain and the United States (Darroch et al. 2001) found female coital experience before age 18 ranging from 50.1 percent for France to 65.2 percent for Sweden.

In Asia, national-level surveys usually find that premarital sex is less widespread than in the West, although smaller in-depth studies focusing on young people’s sexual and reproductive health have revealed that their sexual activity has clearly increased (Jejeebhoy et al. 2001; Gubhaju 2002; Bearinger et al. 2007). Results of surveys of adolescent sexual behaviour suggest that non-trivial percentages of adolescents are now sexually experienced. For example, in the Republic of Korea more than a decade ago, research showed that 11 percent of female and 24 percent of male secondary school students reported having had premarital coital experience (Gayun 1996). A majority of females had had their sexual debut with a boyfriend they planned to marry, while more than half of males had had their first sex with a casual female friend or a commercial sex worker. In a survey undertaken in 2000 in Phnom Penh among rural-urban migrant women aged 15-24 working in the garment manufacturing industry (Nishigaya 2006), a high proportion (54.7 percent) reported having engaged in premarital sex. This study used peer interviewers who were garment workers themselves, and argues that the manner in which sensitive data such as these are collected is crucial to obtaining meaningful results in a society whose culture places considerable weight on female premarital chastity. In the Lao People’s Democratic Republic, a localized survey among community members in Vientiane revealed that sex and pregnancy before marriage were common and more or less accepted because of a common belief that pregnancy outside marriage leads to marriage (Senanikhom et al. 2000).

Sexuality in Thailand has traditionally been a private matter, but the advent of HIV/AIDS in the 1990s opened it to public discussion. A number of studies of adolescent sexual activity have been conducted in Thailand during the last two decades, producing evidence of increasing sexual activity among young people. They have also noted large disparities in rates of sexual activity between young men and young women, although some of this difference may be attributable to female under-reporting and possibly also to male over-reporting (e.g., Koetsawang 1987; Chompootaweep et al. 1991; Rugpao 1997; van Griensven et al. 2001; O-Prasertsawat and Petchum 2004; Srisuriyawet 2006).

Several research teams have studied Thai military conscripts in order to describe the sexual behaviour of the general population of young Thai men. Thai men aged 20-22 who are not in higher education are conscripted by lottery into the Royal Thai Army for two years. Random samples of conscripts therefore provide an excellent representation of men in the lower socio-economic strata of Thai society (Beyrer et al. 1995; Nelson et al. 1996; Celentano et al. 1998). In a study of conscripts from Northern Thailand by Nopkesorn et al (1993), 97 percent reported having had coital experience, with more than half of these reporting having first had intercourse by the age of 16. First sexual intercourse for 74 percent of the sexually experienced men was with a female sex worker, compared to 12 percent with a lover and 8 percent with a female friend. Fully 90 percent of the sexually experienced men had ever had sex with a female sex worker, mostly starting between the ages of 15 and 18. By the age of 16, about half the sample had made their first visit to a female sex worker. Justified as a way of protecting virtuous Thai women from surrendering their virginity, commercial sex had been the major sexual outlet for young unmarried Thai men until AIDS became prevalent in the 1990s (Knodel 1996; Kilmarx et al. 2000).

This picture is, however, changing among a new generation of Thais for several reasons. Taywaditep et al (2004) noted that young women tended to engage in premarital sexual activity more than in previous generations. As perceived and interpreted by Thai people, Western culture has been partially blamed for this change. More recently the transition has also been attributed to men’s heightened fear of HIV. As prevention campaigns have publicized high rates of HIV infection among female sex workers, Thai men have become more cautious over visiting brothel-based sex workers. A decrease in patronage of commercial sex workers among Northern Thai conscripts has been documented over the few years prior to 1996 (Nelson et al. 1996; Celentano et al. 1998).

Nowadays a growing proportion of young men have sex with their girlfriends in the context of committed romantic relationships (Taywaditep et al. 2004). Helped by the anonymity of big cities, the number of cohabitating couples is increasing, much to the disappointment of conservatives concerned for the virtue of Thai women. Although this phenomenon has been inconsistently observed in recent studies, many scholars feel that there is much resistance to it from the Thai public, who are not yet ready to formally approve of these unmarried, sexually active relationships (Soonthorndhada 2002; Jenkins and Kim 2004; Rasamimari et al. 2008).

The literature just reviewed reveals a rapidly changing picture of sexuality and gender in Thailand. In this paper we add to that literature by providing a snapshot of contemporary young Thai people whose sexual practices are overtaking social conventions and thus contributing to poor health outcomes. We report on our recent multi-method study of the sexual experiences of in-school and out-of-school 17-20 year-olds in Chiang Mai City, Northern Thailand. The study combines insights from both quantitative survey data and qualitative data obtained from in-depth interviews and focus group discussions conducted in the Chiang Mai urban area.

The second largest urban centre in Thailand, Chiang Mai City is a major in-migration destination for young people from surrounding areas (Morrison 2004; Vuttanont et al. 2006). We use quantitative data to describe the extent to which male and female respondents from different educational groups have sexual contact or intercourse. The sexual practices of those who had had intercourse (sex) are analyzed, focusing on first sex, most recent sex, casual sex, paid sex, coercive sex and condom use during different kinds of sexual activity. As well, by using qualitative information, we explore the circumstances surrounding intercourse debut, experiences of sexual relationships, and issues bearing on unprotected sexual activity.

Methods

Recruiting survey respondents

Data were collected in 2006. Respondents were recruited from across the range of socio-economic backgrounds represented among adolescents in Chiang Mai. They were males and females aged 17-20 years, both in and out of school or university. Among those who were not studying full-time both employed and unemployed individuals were chosen. We recruited our samples from three sources. The first source was youth-frequented public spaces (assisted by non-governmental organizations (NGOs) that work with Thai youth), from which those sampled fell into three educational groups (out-of-school, vocational school and general school-university). The second source was non-formal education centres, from which all those sampled fell into the out-of-school education group. The third source was formal education centres (vocational schools, senior high schools and university). Thus all sampled persons fell into six groups: males and females who were out-of-school (Group O in tables), studying at vocational schools (Group V), or studying at general high schools or university (Group G). For each group samples were obtained as follows.

The out-of-school sample

Chiang Mai is home to 22 non-formal education centres offering three-hour weekend tutorials to young people not otherwise engaged in education. The six largest centres were chosen and all age–eligible youth present on a teaching day were invited to participate. These young people were outside the formal education system and predominantly engaged in employment, though sometimes unemployed. Respondents were also recruited through four youth-focused NGOs: the Harm Reduction Youth Program, the Street Youth Outreach Team, the Adolescent Sex Education Outreach Team, and the Men’s Sexual Health Outreach Team. Working with NGO staff the field research team (the first author and five Northern Thai research assistants aged under 24 – three females and two males) recruited respondents from an array of public gathering places, including playing fields, shopping malls and public gardens at various times of the day and night. This sample was non-random, but it did ensure that those respondents in the out-of-school group were quite diverse and not dominated by any single source. Overall 132 out-of-school respondents (47 males and 85 females) were recruited from non-formal education centres and 275 (191 males and 84 females) from public gathering places.

The vocational school sample

One public technical, one private technical and one private commercial school were randomly selected from among the two public and 10 private vocational schools in Chiang Mai. Students of target age (17-20) studying electronics, mechanics or computer technology in the technical schools, and marketing, hotel management or finance in the commercial school were invited to volunteer to participate after having the survey and its purpose explained in classroom settings. This yielded 288 private school (133 male and 155 female) and 241 public school (224 male and 17 female) respondents, to whom were added 92 respondents (52 male and 40 female) recruited through public places.

The general school/university sample

From nine public and 11 private high-schools in Chiang Mai one large private and one large public school were selected. Respondents (98 males and 164 females) were drawn from two classes each of Grade 12 students studying pure science, applied science and languages who were invited to volunteer after having the survey and its purposes explained to them. There are also two public and two private universities in Chiang Mai, from which Chiang Mai University was chosen. Respondents (101 males and 226 females) were recruited voluntarily from eight faculties via posters displayed in libraries: the faculties of Medicine, Nursing, Engineering, Agriculture, Humanities, Economics, Accounting and Social Sciences. To these two groups of respondents were added 63 males and 69 females recruited through public places who attended a general school or university.

Sample size targets were based on an error of ±7% at the p<0.05 significance level when measuring key indicators (e.g., the percentage with experience of sexual intercourse). The overall sample was conceptualized as comprising six subsamples representing males and females separately in each educational setting. A required subsample size of 192 was obtained for an expected prevalence of 50 percent (the proportion requiring the largest sample to be estimated with a given level of precision), so that the total sample required was 1152 (6 × 192). In all education and sex groups we exceeded the required sample size except for out-of-school females, who proved more difficult to recruit, and of whom we recruited only 169 (12% less than the target of 192). The final total sample was 1749, including 909 males and 840 females.

Data collection methods

The field research team (age-sex composition and ethnicity noted above) comprised university graduates in sociology or anthropology who were trained in quantitative and qualitative research methods. The sensitive nature of the subject matter dealt with required ethical clearance from both the Australian National University and Chiang Mai University, where the fieldwork was based. It is neither known nor knowable to what extent Thai cultural beliefs that open discussion of sexual matters is inappropriate for young women dissuaded their participation. However, these beliefs are likely to be held more strongly by older generations than targeted here. The first author’s impression is that efforts made to explain the content and purpose of the study (a written information sheet was given to each potential participant), reassurance as to confidentiality, and stressing the right to withdraw at any time led to little difficulty being experienced recruiting female respondents. There were no refusals among young women individually asked to answer the questionnaire (as opposed to being issued a general invitation with classmates or required to respond to a poster in a university library), and no requests to submit to in-depth interview or join a focus group were declined. No instances of severe distress while participating in the project were detected by the research team members.

Quantitative data

Two methods were used to administer the questionnaire. At large schools and Chiang Mai University, with computers and the internet available, respondents completed an online computer-assisted self-interview (CASI – 700 respondents). At other locations a self-administered paper questionnaire (SAQ – 1049 respondents) was used. The questionnaire was 22 pages long and covered socio-economic background, recreational activities, alcohol, tobacco and drug use, relationships, sexual identity and experience, sexually transmitted diseases (STDs), pregnancy, abortion and birth control, mental health, and need for sexual health services. Most questions were pre-coded. Pre-testing of both SAQ and CASI formats was conducted before finalization, and comparison of results obtained from 218 heterosexual male vocational school students who used CASI and 140 who used SAQ provided good assurance that they produced similar results. On 10 behavioural items only one significant difference was found (SAQ respondents had more often smoked during the past year), and proportions acknowledging nominated behaviours were often remarkably similar (e.g., ‘Ever had sexual intercourse’ – CASI 68.8 percent, SAQ 69.2 percent).

Qualitative data

To obtain qualitative data to elaborate on questionnaire findings in-depth interviews and focus group discussions were also conducted, largely with survey respondents but with some focus group members also drawn from outside those ranks. In-depth interviewees were recruited from members of all three educational groups who had completed the questionnaire. Those who appeared interested and willing were invited to participate in the qualitative component. They were recruited steadily throughout the quantitative survey until we reached a point where we figured we had enough. Representation of minority groups in terms of their sexuality was also looked for. Each participant was interviewed by a member of the field research team of similar gender identity for 60-90 minutes. Thirty interviews were conducted (16 with males and 14 with females), with 21 interviewees (11 males and 10 females) being students and nine (five males and four females – six employed and three unemployed) being out-of-school. About half the interviewees lived with one or both parents.

Separate male and female focus groups were conducted, each group comprising 4-8 people and discussions lasting 60-90 minutes. They were facilitated by members of the field research team who had similar gender identities to group members. They took place at a variety of locations: private meeting rooms, or quiet places at playing fields, coffee shops and public gardens. In all there were 16 groups, eight of each sex. Four were recruited largely from the out-of school sample, six largely from the vocational school sample, and six mostly from the general school/university sample. Nine groups (including all four out-of-school groups) were exclusively heterosexual, five had exclusively minority sexuality group memberships (non-heterosexual and questioning), and two had mixed memberships. More than three-quarters of discussants were also survey respondents. Those who were not were generally friends of survey respondents present when focus groups were being arranged, or who accompanied them to focus groups. The group interaction process was used to probe normative aspects of the daily lives and intimate relationships of young people.

Data analysis

SAQ data were double-entered using Microsoft Access 2003. CASI data were digitized as they were collected and merged with SAQ data before analysis using SPSS version 14. Information from in-depth interviews and focus group discussions was collected on digital recorders and in field notes. Digital recordings were fully transcribed in Thai into Microsoft Word documents. The qualitative software package ‘Atlas Ti’ (version 5.2) was used to manage the process of identifying and collecting together data pertaining to repeated normative themes, some of which arose spontaneously from interview interaction and some in response to open-ended questioning about the values, attitudes and practices of young people concerning sexual issues and their reproductive health-seeking strategies. Passages most relevant to the study were later translated into English.

Results

Sexual experience

Sexual contact was defined as having one’s genitals touched by someone else or touching someone else’s genitals for erotic stimulation, including oral sex. Sexual intercourse entailed either vaginal or anal penetration by a male’s penis. Males and females from different educational groups reported different sexual experiences, but sexual contact alone was uncommon. Terms, such as ‘sexual debut’ and ‘first sex’ used hereafter thus refer to the first experience of intercourse, except in female same-sex relationships where penile penetration is obviously not involved. Around a third of all males had never had either sexual contact or sexual intercourse, while 1.9 percent had had sexual contact only, and just under two-thirds had had intercourse (Table 1). By contrast, almost two-thirds of female respondents had never had either sexual contact or sexual intercourse, 2.3 percent had had sexual contact only, and a third had had sexual intercourse.

Table 1.

Sexual experience of adolescents by sex and educational group.

	Ever had sexual contact or sexual intercourse?
Sex and Educational Group	Never had either		Had sexual contact only		Had sexual intercourse		Total
Males	No.	%	No.	%	No.	%	No.	%
Out-of-school	45	18.9	0	0.0	193	81.1	238	100
Vocational school	116	28.6	11	2.7	279	68.7	406	100
General school and university	163	62.2	6	2.3	93	35.5	262	100
Total	324	35.8	17	1.9	565	62.4	906	100
Females	No.	%	No.	%	No.	%	No.	%
Out-of-school	76	45.0	4	2.4	89	52.7	169	100
Vocational school	73	34.6	6	2.8	132	62.6	211	100
General school and university	382	83.2	9	2.0	68	14.8	459	100
Total	531	63.3	19	2.3	289	34.4	839	100

Eight in ten out-of-school males and over two-thirds of vocational school males had had sexual intercourse (Table 1). Almost two-thirds of vocational school females also reported having had intercourse, and this was easily the group for which the gender difference in intercourse experience was smallest. Out-of-school females recorded a 53 percent intercourse prevalence figure, reflecting in part a subsample of 27 who were of Hill Tribe ethnicity and Christian religion. Recruited at non-formal education centres, they occupied boarding houses run by Catholic nuns, and none had had intercourse. By contrast, of 30 out-of-school females who were Buddhist, Northern Thai and not non-formal education students, 26 (87 percent) had had intercourse (p<.001). If the subsample of 27 is discounted, intercourse experience among out-of-school females rises to 63 percent, almost identical to the figure for vocational school females. Males and females from the general school and university group were the least sexually experienced, only 35.5 percent of males and 14.8 percent of females having had intercourse.

Numbers of partners

Among those with intercourse experience the mean number of lifetime sexual partners was 6.6 (median 4.0) for males and 3.5 (median 2.0) for females (p<.001). The mean number of sexual partners for out-of-school males was 8.4 (median 5.0), while it was 6.2 (median 3.0) for vocational school males and 4.2 (median 2.0) for general school and university males (p<.05). Among females, the mean numbers of sexual partners were 5.1 (median 3.0) for the out-of-school group, 3.2 (median 2.0) for the vocational school group and 2.0 (median 1.0) for the general school and university group (p<.05).

First sexual intercourse

Sexually experienced respondents reported first intercourse at a mean age of 16.73 years. Sexual debut had mostly occurred in their room/house, a partner’s room/house, or at a friend’s place (Table 2), and the first sexual partner was usually a girlfriend/boyfriend (81.6 percent). Very few males (0.9 percent) said their first sexual partner was a sex worker, consistent with evidence of mass abandonment of sexual initiation with sex workers in the wake of the HIV/AIDS epidemic (Nelson et al. 1996; Celentano et al. 1998). Condoms were the most popular method of contraception at sexual debut, but less than half (45.0 percent) had used them; others had used withdrawal (30.3 percent) or the morning after pill (4.6 percent). Males were generally more likely to report condom use and females use of withdrawal. Over 18 percent of respondents had not used a contraceptive method at first intercourse, with out-of-school females most likely to be in this category and general school/university females least likely to be.

Table 2.

Details of first coitus of adolescents who had had sexual intercourse by sex and educational group.

	All	Males				Females
	Total (n=850)	Group Oa (n=193)	Group V (n=275)	Group G (n=93)	Total (n=561)	Group O (n=89)	Group V (n=132)	Group G (n=68)	Total (n=289)	Total (standardized)^
	%	%	%	%	%	%	%	%	%	%
Where it happened *#
Own room/house	37.1	46.1	45.8	39.8	44.9	19.1	23.5	22.1	21.8	21.8
Partner’s room/house	34.2	18.7	26.2	23.7	23.2	48.3	60.6	55.9	55.7	55.6
Friend’s room/house	20.5	26.9	22.2	18.3	23.2	23.6	10.6	13.2	15.2	15.5
Motel or hotel	5.2	4.1	3.3	12.9	5.2	7.9	3.8	4.4	5.2	5.3
Car	0.8	0.5	0.0	3.2	0.7	0.0	0.8	2.9	1.0	0.9
Other	2.2	3.6	2.6	2.2	2.8	1.1	0.8	1.5	1.0	1.0
First partner *+
Boyfriend/girlfriend	81.6	75.1	81.5	73.1	77.9	79.8	93.2	91.2	88.6	88.3
Other friend	6.3	10.9	5.4	8.6	7.8	2.2	3.0	5.9	3.5	3.2
Acquaintance	8.0	9.8	8.0	11.8	9.3	12.4	3.0	1.5	5.5	6.0
Stranger	2.5	2.1	3.6	4.3	3.2	1.1	0.8	1.5	1.0	1.0
Sex worker	0.6	1.6	0.4	1.1	0.9	0.0	0.0	0.0	0.0	0.0
Other	1.1	0.5	1.1	1.1	0.9	4.5	0.0	0.0	1.4	1.5
Sex of first partner *#
Male	38.0	3.7	7.2	25.8	9.1	97.8	90.9	95.6	94.1	94.1
Female	62.0	96.3	92.8	74.2	90.9	2.2	9.1	4.4	5.9	5.9
Method used to avoid disease/ pregnancy at first intercourse *+
Withdrawal	30.3	18.1	31.9	21.5	25.4	27.0	52.3	32.4	39.8	40.3
Condom	45.0	59.1	46.7	52.7	52.0	32.6	22.7	47.1	31.5	30.2
Morning after pill	4.6	3.1	4.3	5.4	4.1	7.9	3.0	7.4	5.5	5.4
Other	1.6	1.0	0.8	1.1	0.9	2.2	3.0	3.0	2.7	2.7
No method	18.6	18.7	16.3	19.4	17.6	30.3	18.9	10.3	20.4	21.4

^aGroup O=Out-of-school; Group V=Vocational school; Group G=General school/university

^*Chi-square statistic between males and females significant at 0.05 level.

^#Chi-square statistic between educational groups of males significant at 0.05 level.

⁺Chi-square statistic between educational groups of females significant at 0.05 level.

^{^}Standardized to male distribution by educational group.

Males reported slightly younger ages at sexual debut than females, with a mean age of 16.55 years compared to 17.03 years (p<.001). Out-of school males had the youngest mean age at debut of 15.97 years, as against 16.85 and 16.84 years for vocational school and general school/university males respectively (p<.001). Out-of-school females also had the youngest mean age at first intercourse of 16.48 years, compared to 17.24 and 17.35 years for vocational school and general school/university females (p<.001).

Males, particularly from vocational schools and out-of-school, tended to report that first intercourse had occurred in their own room/house while females, especially from vocational schools, agreed that it had mostly happened at their partner’s place (Table 2). More females than males described their first sexual partner as a boyfriend/girlfriend while more males claimed that person was an acquaintance, other friend or stranger. Out-of-school females were much more likely than other females to report having first had sex with an acquaintance. Nearly 10 percent of sexually experienced males reported that their first sexual partner was of the same sex, compared to 5.9 percent of females.

Circumstances surrounding sexual intercourse debut

Among males first intercourse was often unplanned; an unpremeditated sexual relationship had developed, with alcohol sometimes being involved. Others explained that it had occurred in a dating context, where a sexual relationship had developed more gradually.

The first woman in my life that I slept with was my ex-girlfriend. I met her a couple of years ago and we saw each other for many months before I had sex with her. It happened after we went out to a party. She could not go home because it was too late, so she stayed overnight in my friend’s room. I did not plan to have sex with her, but it just happened. (Toey, male, 19 years, technical school student)

Some males even stated frankly that they had deliberately sought out sexual experience, particularly after watching pornography or listening to accounts of the sexual exploits of male friends.

I can say that most boys in this factory have had the experience of watching porn. They usually watch with a group of friends and discuss whether the girls act well in bed, and sometimes they share the experiences of having sex. … For me, my first sex occurred a year ago. One night after watching a movie at the cinema, I took her to my place. We talked about things, including the movies and sex scenes. And then having sex with her just happened. (Tai, male, 19 years, factory worker)

Girls tended to portray their first sexual experiences more sentimentally. By and large they had engaged in sexual relations to display love or cement committed relationships.

My first sex happened simply because of love. I love him [her boyfriend] and he loves me. That’s all. This is the most important reason why I went out with him. (Duan, female, 18 years, general school student)

Some also reported a need to satisfy their inquisitiveness, stating that they wanted to conduct an experiment.

I think, sometimes, it is worthwhile to try [having first sex] when we feel like that one is the right person. (Lek, female, 20 years, commercial school student)

Most recent sex

As Table 3 shows, more than half the sexually experienced respondents had had intercourse within the last week, usually with their boyfriend/girlfriend. Some males in particular did, however, report the latest partner to be an acquaintance, other friend or stranger. Fewer than half the experienced respondents reported condom use at most recent intercourse. Others reported using withdrawal, the pill and the morning after pill, but 17.9 percent had used no method of contraception. Out-of-schoolers of both sexes were the most likely to have used no method; those attending general schools or university the least likely.

Table 3.

Details of most recent coitus of adolescents who had had sexual intercourse by sex and educational group.

	All	Males				Females
	Total (n=850)	Group Oa (n=193)	Group V (n=275)	Group G (n=93)	Total (n=561)	Group O (n=89)	Group V (n=132)	Group G (n=68)	Total (n=289)	Total (standardized) ^
	%	%	%	%	%	%	%	%	%	%
Timing of most recent intercourse +
Within last two days	32.5	33.7	29.8	21.5	29.8	51.7	35.6	23.5	37.7	39.1
Within last week	24.4	24.4	23.3	29.0	24.6	20.2	28.0	20.6	23.9	24.1
Within last month	17.1	13.0	20.4	14.0	16.8	12.4	16.7	26.5	17.6	16.8
Within last 3 months	13.3	16.6	14.5	14.0	15.2	5.6	10.6	13.2	9.7	9.3
Within last year	8.2	9.8	7.3	12.9	9.1	7.9	4.5	8.8	6.6	6.4
More than a year ago	4.6	2.6	4.7	8.6	4.6	2.2	4.5	7.4	4.5	4.2
Most recent sex partner *
Boyfriend/girlfriend	85.9	80.3	81.8	79.6	80.9	94.4	96.2	95.6	95.5	95.5
Other friend	3.8	5.7	5.1	3.2	5.0	1.1	0.8	2.9	1.4	1.3
Acquaintance	6.7	7.3	9.9	12.9	9.5	2.2	0.8	1.5	1.4	1.4
Stranger	2.2	4.7	2.2	2.2	3.0	0.0	1.5	0.0	0.7	0.7
Sex worker	0.6	1.6	0.7	0.0	0.9	0.0	0.0	0.0	0.0	0.0
Other	0.8	0.5	0.4	2.2	0.7	2.2	0.8	0.0	1.0	1.2
Sex of most recent partner *#
Male	38.3	3.6	8.7	28.0	10.2	96.6	89.4	94.1	92.7	92.7
Female	61.7	96.4	91.3	72.0	89.8	3.4	10.6	5.9	7.3	7.3
Method used to avoid disease/pregnancy at most recent intercourse *#+
Withdrawal	28.9	17.6	33.2	28.0	27.0	19.1	42.7	30.9	32.6	32.6
Condom	41.2	53.4	41.2	52.7	47.3	34.8	18.3	42.6	29.2	28.0
Oral pill	7.9	4.1	7.7	2.2	5.5	14.6	14.5	5.9	12.5	13.1
Morning after pill	2.2	1.0	1.5	4.3	1.8	0.0	3.1	7.4	3.1	2.7
Other	1.9	2.1	1.1	1.1	1.4	2.2	3.1	3.0	2.7	2.7
No method	17.9	21.8	15.3	11.8	17.0	29.2	18.3	10.3	19.8	20.7

^aGroup O=Out-of-school; Group V=Vocational school; Group G=General school/university

^*Chi-square statistic between males and females significant at 0.05 level.

^#Chi-square statistic between educational groups of males significant at 0.05 level.

⁺Chi-square statistic between educational groups of females significant at 0.05 level.

^{^}Standardized to male distribution by educational group.

More sexually experienced females than males reported having sex in the last week, and more than half of out-of-school females reported having it within the last two days, well above figures for other sex-education groups. More females than males reported a boyfriend/girlfriend being their most recent partner, and few most recent partners of males had been sex workers. Males, especially general school/university students, reported their most recent sexual encounter being a same sex encounter slightly more often than females did.

Males reported much higher condom use at last coitus. Withdrawal was the most common contraceptive method reported by females, who also more often reported using the pill at last sex. Vocational school males and females had most often used withdrawal, but were least likely to have used condoms. Those from the general school/university group tended to use condoms, with out-of-school males also claiming high condom use. Unfortunately we do not have direct information connecting contraceptive use at the most recent sexual encounter and type of relationship, but we do know that 80-90 percent of relationships were boyfriend-girlfriend.

Casual sex

Around a third of sexually experienced respondents had had casual sexual partners (Table 4). Over half claimed a condom had always been used in casual encounters. Males were far more likely than females to claim casual experience, and there was a statistically significant gender difference in the frequency with which condoms were claimed to have been used in casual sexual encounters – 57 percent of males claimed use at every such encounter compared to only 36 percent of females after standardization for educational group. The gender difference in reported regular condom use within the general school/university group is huge, but only seven females in this group conceded experience of casual sex.

Table 4.

Casual coital experiences of adolescents who had had sexual intercourse by sex and educational group.

	All	Males				Females
	Total (n=850)	Group Oa (n=193)	Group V (n=275)	Group G (n=93)	Total (n=561)	Group O (n=89)	Group V (n=132)	Group G (n=68)	Total (n=289)	Total (standardized) ^
	%	%	%	%	%	%	%	%	%	%
Ever had a casual sexual partner *#+	34.2	50.8	40.5	34.4	43.0	31.5	10.2	10.3	16.9	17.5
Number of casual coital encounters	(n=291)	(n=98)	(n=112)	(n=32)	(n=242)	(n=29)	(n=13)	(n=7)	(n=49)
Only one	26.6	23.5	25.9	31.3	25.6	32.1	23.1	42.9	31.3	29.5
2-5	47.6	45.9	50.0	43.8	47.5	46.4	53.8	42.9	47.9	49.4
6-10	8.3	9.2	5.4	12.5	7.9	10.7	15.4	0.0	10.4	11.2
More than 10	17.6	21.4	18.8	12.5	19.0	10.7	7.7	14.3	10.4	9.8
Frequency of condom use with casual partners *
Always	53.8	57.1	54.5	65.6	57.0	42.9	38.5	14.3	37.5	36.0
Often	15.2	16.3	11.6	6.3	12.8	25.0	38.5	14.3	27.1	29.8
Sometimes	16.6	13.3	20.5	12.5	16.5	17.9	7.7	28.6	16.7	14.7
Never	14.5	13.3	13.4	15.6	13.6	14.3	15.4	42.9	18.8	19.6

^aGroup O=Out-of-school; Group V=Vocational school; Group G=General school/university

^*Chi-square statistic between males and females significant at 0.05 level.

^#Chi-square statistic between educational groups of males significant at 0.05 level.

⁺Chi-square statistic between educational groups of females significant at 0.05 level.

^{^}Standardized to male distribution by educational group

Experiences of sexual relationships

Female in-depth and focus group informants, especially, expressed a range of feelings after their first experiences of sexual intercourse. Although first sex could be exciting, many disclosed feelings of anxiety, shame and guilt when recalling it. Feelings about subsequent sexual intercourse episodes were also varied, ranging across reluctance, embarrassment, excitement, fun and boredom.

After I slept with him [ex-boyfriend] I was really scared that my mum would find out. I cried alone at home as I knew what I had done was wrong. I know that a proper girl should not do that. (Duan, female, 18 years, general school student)

I felt ashamed of myself when I had sex with my first boyfriend. Girls have never been taught that having sex before they are married is acceptable. (Jeeb, female, 18 years, commercial school student)

Serial monogamous relationships were a pattern among some males and females. Many males, especially, believed that love did not last forever. If the current relationship broke up they could enter another. Some female informants argued that they did not need to reveal past sexual experiences to a new partner. They preferred to conceal them.

I had sex with only two girls, one at a time. After I broke up with my first girlfriend it took me a few months to meet the second. Now I am still seeing her. We get along well. … I don’t think I want to sleep with other girls. My girlfriend would kill me if she found out. (Toey, male, 19 years, technical school student)

In my life I have had three men. When my first boyfriend left me, it really hurt. Then I met the second. We had a very good time for months and then he moved to study in Bangkok. We couldn’t meet much. Anyway, finally I met the third guy. He lives in Chiang Mai and we are so close. So I decided to break up with my second and go out with my [current] boyfriend. … I don’t need to let him know all of my past. Just tell him only what I think he can accept. (Fresh, female, 19 years, commercial school student)

However, non-monogamous sexual relationships (both cheating and consensual non-monogamy) also existed among some young people, particularly out-of-schoolers. Some males claimed concurrent intimate relationships, and while they may be uncommon, some females were encountered who besides having boyfriends also had sexual relationships with middle-aged married men. They had sex with these men to express their gratitude for money or other support received from them, but usually kept these relationships secret from boyfriends.

I know that I have him [a boyfriend], but I still got involved with another man. It’s quite complicated. He [the boyfriend] sometimes spends too much time with his mother and ignores me. … As I work as a ‘Cheer Beer’ girl [selling alcohol in a pub or restaurant], I have a lot of chances to see [married older] men. I met a man who is very nice to me. He offers lots of stuff to me and supports me when in need. So, I slept with him. … Two months ago it was my boyfriend’s birthday. I also got some money from that man to buy a very nice gift for my boyfriend. (Noi, female, 20 years, pub employee)

Overall, 7.8 percent of sexually experienced respondents had accepted money and/or other resources to allow someone to have intercourse with them, while 5.0 percent had ever paid someone else for sex. Consistent condom use when having paid sex was reported by 60.7 percent of those who had accepted payment and/or paid for sex. Males were more likely to have paid (8%) and females more likely to have been paid (7%) for sex. However, claimed consistent condom use during paid sex was higher among males. A quarter of females who had engaged in paid sex reported never using condoms while doing so, compared to 7.7 percent of males, with no significant difference by educational group.

Issues bearing on protected and unprotected sexual activity

Sexually experienced informants, regardless of gender and sexual preference, often used a similar logic. Whether they engaged in protected or unprotected sex depended on who they had sex with, the timing of the encounter, and whether condoms or other contraceptives were available. Many knew well the potential negative consequences of unprotected sex, but were unlikely to have serious concerns provided partners were not sex workers. The common logic was that if they had any form of casual or paid sex they preferred to use condoms, if available, to avoid sexually transmitted infections. But if having sex with loved ones or other trusted sexual partners they preferred other contraceptive methods to avoid pregnancy. They equated being monogamous with being safe, and believed that if sure their partner was not promiscuous, condoms were unnecessary. Some reasoned that their partner was a virgin when they met; hence protection to avoid disease was not needed. Others stated that loyalty and mutual intimacy created trust, and ensured there was no risk of STD or HIV infection.

We know that using condoms can protect us from getting AIDS or STDs. If we think we are at risk, we should use them …. Everyone knows that. But at the time of having sex, we kind of forget everything (laughing). (Male focus group discussant, 18-20 years, technical school)

We are told condoms are good for protection. But I don’t see many boys carrying them. Maybe it’s hard to find free condoms, or the boys can’t afford to buy them. … I think condom use is good in theory, but in the real world I can guarantee that no one uses them all the time. That’s impossible. (Female focus group discussant, 18-20 years, neighbourhood community)

Young people are slow to adopt contraceptives and use them irregularly. Many informants had used at least one contraceptive method, but use was infrequent. Regularly, sex was unprepared for, unexpected, and therefore unprotected. Apart from condoms, use of several different kinds of conventional contraceptive methods was reported, including withdrawal, pills and emergency contraceptive pills. Many had used a ‘safe period’ technique to avoid pregnancy, although they were likely to lack accurate knowledge of what the safe period was. Peers, print media and the internet were the main sources of contraceptive information.

My boyfriend used a condom with me only once. Other than that we use withdrawal and safe period techniques. … Well, as in my understanding, to count the safe period is … it is seven days before the period and seven days after the mens[es] stop. Is that right? (Jane, female, 18 years, unemployed)

When asked about sexual responsibility and choices aimed at avoiding disease and unwanted pregnancy, many girls said it was boys’ decision whether to use a condom or a technique like withdrawal, since those were perceived to be men’s methods. Other methods, if needed, were girls’ responsibility to seek out and employ. Most informants knew that condoms were the best method for combined disease and pregnancy prevention, but felt they might not be needed with trusted sexual partners.

I think both boys and girls should be responsible for protecting against any bad outcomes due to having sex …. But, you know, in the real situation I don’t see many boys who are keen to use condoms. They leave this job [of protection] to girls. (Male focus group discussant, 18-20 years, general school)

Young people in our study sample often experienced adverse health outcomes of their ambivalent attitudes to protected sex. For example, among the one-third of females who had had sexual intercourse, 32 percent had become pregnant at least once. Of these 64 percent had had an abortion, 15 percent a miscarriage, and 32 percent delivered a child (see Tangmunkongvorakul et al. 2010 for details). Nine percent of young men and women had contracted STDs such as Gonorrhoea, Syphilis and Chlamydia, but many more reported having experienced symptoms of STDs (Tangmunkongvorakul 2009).

Traditional gender roles that limit negotiation of protected sex

Ambivalent social attitudes towards sexuality have resulted in young people receiving conflicting messages. Traditionally, Thai boys and girls are brought up and disciplined differently, double standards in role expectations and appropriate conduct being imparted from an early age. Interviews with adolescents provided considerable evidence that a gender double standard of sexual morality remains strong in Thai families. Sexual expression and experience were acceptable for young unmarried men, but not for young unmarried women.

Because young women were expected to preserve their innocence and remain naive about sex until marriage, it was difficult for them to have an open discussion of sex or condom use with potential partners. Any deviation from this norm on the part of young women was considered by parents and families to be detrimental to their reputations and perhaps to their marriage prospects. Sexual matters were not commonly discussed between children, especially daughters, and parents, the latter fearing this might lead to sexual experimentation.

At home when I talk to my parents, they usually ask me if I have a boyfriend at Uni. They are very worried about this. I said no. … They may be afraid that if I had a boyfriend, I might do something wrong [have sex] with him. My parents don’t talk about sex issues with me directly. (Pim, female, 19 years, university student)

Gender double standards and power relations also persist in youthful sexual partnerships. Interviews showed that when young men and women first entered close relationships, women normally felt they had similar power to men. They could make choices about how intimate the relationship became. However, after engaging in sexual intercourse, some women perceived that they had less power to negotiate sexual issues than men. Perceptions of female informants about love largely involved, on the one hand, a wish to accede to a boyfriend’s desire for sex, and on the other, fear that reluctance to engage in sex, or bringing up problems that might arise from sex, would be met with indifference at best and abandonment at worst. The qualitative data yielded evidence confirming that girls often were unable to negotiate protected sex, and boys failed to fully appreciate the potential negative consequences of their sexual behaviour.

Talking about a serious issue like this [protected sex] is very hard for me. I had a lesson with my ex-boyfriend. I really wanted him to use condoms. I talked to him frankly, but he didn’t say anything. His reaction was very disappointing. When I said no to having sex when he didn’t bring a condom, he really got in a bad mood. It was a very uncomfortable situation. … So with my current boyfriend, if he doesn’t bring a condom we use some other method. I don’t want to be fussy about this any more. (Nook, female, 20 years, university student)

Coercive sex

Sexual coercion ranges from forceful rape to nonphysical forms of pressure that compel people to engage in sex against their will. Table 5 shows that 14.7 percent of sexually experienced respondents had ever been forced, physically or psychologically, to have intercourse. Boyfriends/girlfriends were the main perpetrators followed by acquaintances, other friends and strangers. The majority of those reporting coercion said it had occurred the first time they had ever had sexual intercourse. Half of them had used a condom on the occasion of coerced first sexual intercourse. This is not surprising, as willingness to use a condom could be a common strategy in securing cooperation to engage in psychologically coerced intercourse.

Table 5.

Coercive sex experience of adolescents who had had sexual intercourse by sex and educational group.

	All	Males				Females
	Total (n=850)	Group Oa (n=193)	Group V (n=275)	Group G (n=93)	Total (n=561)	Group O (n=89)	Group V (n=132)	Group G (n=68)	Total (n=289)	Total (standardized) ^
	%	%	%	%	%	%	%	%	%	%
Ever forced physically or psychologically to have intercourse against will *	14.7	10.9	9.1	3.2	8.8	33.7	22.0	25.0	26.4	26.5
Sex partner first time had intercourse against will +	(n=124)	(n=21)	(n=25)	(n=3)	(n=49)	(n=30)	(n=28)	(n=17)	(n=75)
Boyfriend/girlfriend	61.0	52.4	58.3	66.7	56.3	43.3	82.1	70.6	64.0	66.8
Other friend	11.4	19.0	12.5	0.0	14.6	13.3	0.0	17.6	9.3	7.5
Acquaintance	17.9	14.3	16.7	0.0	14.6	36.7	7.1	11.8	20.0	18.1
Stranger	6.5	4.8	12.5	33.3	10.4	3.3	7.1	0.0	4.0	4.6
Other	3.3	9.5	0.0	0.0	4.2	3.3	3.6	0.0	2.7	2.9
Was first time ever had intercourse	67.7	66.7	60.0	66.7	63.3	63.3	82.1	64.7	70.7	72.7
Condom used on that occasion	49.2	57.1	64.0	66.7	61.2	56.7	28.6	35.3	41.3	39.4

^aGroup O=Out-of-school; Group V=Vocational school; Group G=General school/university

^*Chi-square statistic between males and females significant at 0.05 level.

⁺Chi-square statistic between educational groups of females significant at 0.05 level.

^{^}Standardized to male distribution by educational group.

There was a statistically significant gender difference in reporting coercive sex, fewer than one in ten sexually experienced males reporting it compared to more than a quarter of females (and a third of out-of-school females). However, similar proportions of males and females reported that the first coercing partner had been a girlfriend/boyfriend, and that coercion had first happened at first intercourse. Three-quarters of females who were still in school deemed their coercive sexual partner a loved one, but more than a third of coerced out-of-school females classified the sex partner as an acquaintance. There was no significant difference by educational group among males.

Informants spoke of coerced sex in in-depth interviews. Female informants recounted several instances, all at sexual debut and mostly with a person they knew. Coercion ranged from non-physical forms such as verbal abuse and forced viewing of pornography, through unwanted touching or fondling, to attempted and forced penetrative sex.

I am not sure if this is what is called a rape. I first had sex with my [ex-] boyfriend in a public garden. At that time I went out with him and dropped by to his place. We watched a home movie and after that he turned on the porn. I was so embarrassed when watching with him. … After that, at almost midnight, he brought me home, but we stopped by to have some fresh air and sneaked into the public garden. After talking for a while he pushed me on the grass field near a bush and tried to have sex with me. … I was so scared and slapped his face and was crying. He was only saying that he loved me so much. After a while when I was calmed down he started to touch me again. By that time I didn’t have any energy to resist it, and just let him do it. (Fresh, female, 19 years, commercial school student)

Discussion

Although there have been many previous studies focusing on adolescent sexual behaviour in Thailand, most have been based entirely in educational institutions (Koetsawang 1987; Chompootaweep et al. 1991; van Griensven et al. 2001; O-Prasertsawat and Petchum 2004; Srisuriyawet 2006). Only a few have included both males and females from in- and out-of-school settings (Podhisita et al. 2001a), and none has examined differences in intimacy and sexual practice by sex and educational background. The results from this study show that adolescent sexual behaviour varies depending on gender and educational background. Analysis refined by these variables and supported by insights gained from qualitative research helps to present a clearer picture of sexual practices among young Northern Thai people.

The study has drawn together quantitative and qualitative information on sexual practices among young Thais. The results indicate that they frequently engage in, or are subjected to, a variety of behaviours that may expose them to the risk of contracting sexually transmitted infections and experiencing unwanted pregnancies. These include having multiple sexual partners, engaging in unprotected intercourse with various types of sexual partner (steady, casual and paid) and experiencing sexual coercion (Manopaiboon et al. 2003; Liu et al. 2006).

Intimate feelings, sexual curiosity, sexual desire and experimentation are the main reasons young people engage in sexual activity. When a couple is alone, the pressure to do so is intensified. Although some young people, especially females, perceive it to be inappropriate to engage in sexual relationships, curiosity, desire, finding themselves in situations of complete privacy and, on occasion, a partner’s physical or psychological coercion can lead to unplanned sexual relations (Cash et al. 1997; Im-em et al. 2005).

Some young people, young women especially, feel ambivalent about sexual activity. While it is joyful and pleasurable, they may also feel guilt and anxiety. This ambivalence is intensified in modern Thai society because young people have more opportunities in education and the workforce, and feel less controlled by their parents (Soonthorndhada 2002; Thianthai 2004). This makes them feel independent, and entitled to commence sexual activity at whatever age seems appropriate to them.

In rapidly globalising countries like Thailand, young people are at the intersection of influences on sexual practices via the modern media and technology and the cultural traditions of earlier generations. Young Thai women have high status and freedom in comparison to women from countries such as Pakistan (Hamid et al. 2010). Nevertheless, they experience cultural values that inhibit their sexuality outside marriage and open discussion with adults about sexual matters, just like young women from Pakistan and many other globalising countries, such as India, South Africa (Lambert and Wood 2005) and Mongolia (Roberts et al. 2005). Due to a cultural understanding that sex in Thailand is practiced in prescribed circumstances and not discussed, the foundations for modern approaches to sex education are shaky. At the same time Thai society clings to a powerful gender double standard, providing more liberty to men than women and placing more value on Thai women’s virginity. This can create a power imbalance favouring males over females in sexual relationships, causing the latter to internalize a passive and non-initiative-taking role. Framing virginity as a ‘gift’ has been shown to disempower women in other countries (Carpenter 2002). In the era of HIV/AIDS, and due to a well-publicised health campaign, young Thai men are likely to be fully aware of the benefits of, and convinced of the necessity of, protected sex with sex workers, among whom rates of HIV were almost 50 percent in early 1990s (Nelson et al. 1996). They are less convinced of the need for condom use with other women, such as girlfriends. As in Mongolia, traditional gender roles also limit the ability of young women to negotiate safe sex because those young women should not be sexually experienced.

This study found that young people reported quite low levels of condom use when having sex. Although substantial proportions of respondents from vocational schools and out-of-school were sexually active, consistent condom use was low and most respondents did not perceive themselves to be at high risk of acquiring sexually transmitted infections. Results from other research (VanLandingham et al. 1993; Cash et al. 1997; Morrison 2004) have suggested that previous prevention efforts in Thailand have unintentionally conveyed the message that only high risk groups like sex workers and their clients could become infected. It is possible that these campaigns underlie young people’s perceptions of invulnerability to HIV and STD risk. Moreover, the fact that less than half of sexually experienced young people reported using a condom at first intercourse suggests that unprotected first intercourse may prevail among the broader population of Thai young people.

That almost half of sexually active young people were using either an ineffective method of contraception or no method at all at most recent coitus (Table 3) highlights another area of concern. Critiques of family planning services in Thailand indicate that most contraceptive efforts have concentrated on married couples (see, for example, Knodel and Pramualratana 1996). Meanwhile, contraceptive use by unmarried young people has been ignored, because they are not supposed to have sex (Gray 1999). Young people, particularly females, have nevertheless indicated that they would like more access to contraceptive information and services (Soonthorndhada 2002; Tangmunkongvorakul et al. 2005).

The fact that more than a quarter of sexually experienced young women had been either physically or psychologically forced to have intercourse against their will, usually by a loved one or acquaintance, is disturbing. It suggests, as studies elsewhere in Thailand have, that coercive sexual conduct may be a significant cause of exposure to sexual health risks among young Thai women (Cash et al. 1997; Allen et al. 2003; Im-em et al. 2005). Among female respondents who reported coercive sex, on 40 percent of these occasions a condom was used. This high figure suggests that condoms may be employed as tools of persuasion. However, it is a further cause for concern that on the other 60 percent of coerced occasions condom use was not reported, and we do not know if any other methods were used to prevent pregnancy or STIs. Health interventions that seek to improve young women’s sexual health by addressing sexual coercion need to target men as well as women, since if the former did not coerce, the latter would not need to resist (Allen et al. 2003). At present sexual health programs focus on sex education and individual counselling, and there are no interventions designed specifically for young people that address sexual coercion and other sexual health issues.

Results from this study indicate that a variety of sexual risk-taking behaviours take place among Northern Thai young people. Condom use should be a central focus for initiatives to prevent adverse adolescent sexual health outcomes, although new intervention approaches to encourage the use of other contraceptives are also needed. It is clear that intimate sexual partnerships are nowadays commonplace among Thai young people, and that partner turnover is quite frequent, increasing the potential for sexually transmitted infections to spread. This suggests the need to direct more effort toward condom use, as our data suggest they are used by approximately half the respondents during sexual intercourse. As also advocated in previous research (Jenkins et al. 2002), findings here suggest several areas of intervention that should be explored. They include targeted education programs offered in school and non-school settings that develop an understanding of intimate relationships, foster a positive attitude to condom use at first intercourse, relate condom use to other age-appropriate coping skills, enhance future intentions to use condoms, and increase incorporation of condom use into broader birth control efforts. Findings here, as well as those from other research (Podhisita et al. 2001b; Allen et al. 2003; Jenkins and Kim 2004), suggest that changes in sexual norms among young people in Thailand need to be recognized and accepted in order to mount effective programs and interventions to combat negative sexual health consequences. Accepting recent changes in the lifestyles and sexual norms of young Thais has been difficult for older Thai people and institutions (Thianthai 2004; Vuttanont et al. 2006). But the young are living in a new, unprecedented world, and it seems crucial that this happen if new and ongoing interventions are to succeed in promoting adolescent sexual health.

Study limitations

There were several issues encountered when conducting our research. The first and largest was that the topic being investigated is personal and intrinsically sensitive. To address this we used trained field staff who were young themselves and could relate well to the study population. The representativeness of our sample was also a problem, as we were studying a hard to reach population of young adolescents, particularly those who were out-of-school. We chose to pre-stratify by those attributes which were likely to influence the outcomes of interest – especially sexual experience. We also chose a large and diverse sample, which although not random could capture the extant variation credibly and gave us subsamples of the three educational categories for males and females. We have remained aware of the fact that our samples are not random when interpreting our quantitative data, and that this is a limitation of our study. A random sample would, however, have been impracticable.