Introduction

Religion governs human behaviour even as globalisation and modernisation bring changes in people’s lifestyles. To cope with modernisation, some might ignore their traditional beliefs, including religion, while some try to adapt and integrate changes into a modern life. Having a religion does not mean that one is religious. Some might declare themselves as having a religion and admit the presence of God in their life but not practise religious teachings. Religiosity reflects the degree of one’s devotion and passion towards religion (Salleh 2012). It is affected by biological and social factors (Bradshaw and Ellison 2009). Religious activities such as private prayers and reading religious scriptures vary according to age, gender and ethnicity (Burdette and Hill 2009). As age increases, religious activity decreases, especially in male youths (Steinman and Zimmerman 2004). Females with a low socio-economic status have been found to be more religious compared to males (Collett and Lizardo 2009).

Religion brings people together through shared activities, especially within religious congregations (Graham and Haidt 2010). Religious affiliation provides social and emotional support and teaches people to be generous and helpful (Graham and Haidt 2010; Mojahed 2014). Youth who participate in religious activities internalise religious orders and prohibitions and thus perform activities in accordance with the expected norms of their religion and religious comrades (Smith 2003). Further, having a strong personal faith and practising personal prayers have additional influences in shaping people’s behaviour, and these people tend to have fewer behavioural problems (Ahmed et al. 2011). Religious youth are associated with lower chances of risky behaviours, including smoking, alcohol consumption and illicit drug use (Sanchez et al. 2011; Sinha et al. 2007a; Graham and Haidt 2010; Mason et al. 2012). This relationship has also been observed in youth sexual activity. A longitudinal study among youth ages 15–16 in the USA showed that high religiosity delayed sexual intercourse by two years, regardless of gender (Hardy and Raffaelli 2003). Two other studies among Nigerian and US youth suggested a gender difference in the relationship between sexual attitude and initiation, in which the influence of religiosity was greater more on female than male youth (Fatusi and Blum 2008; Ahrold et al. 2011).

Different types of religions and settings can give different meanings to certain behaviours. In a study involving 90 countries, Muslims and Hindus were found to have low rates of premarital sex compared to those of other religions (Adamczyk and Hayes 2012). Religion was influential in the sexual activity of Latina girls in the USA but less determinative among Australian girls (Edwards et al. 2011; Siebold 2011). It is common in Muslim culture for parents to use religious views, such as the prohibition on premarital sexual activity in Islam, to educate their children on sexual issues. Malaysian parents preferred sex education to be delivered from Islamic perspectives (Makol-Abdul et al. 2009). However, once youths make connections with other people, peers might become their most important sex educators (Alquaiz et al. 2012). To youths, conforming to peers’ sexual activity is more important than practicing their religious teaching and activities (Steinman and Zimmerman 2004; Smith 2003). Age, maleness and living away from parents were associated with youth sexual activity (Mudassir et al. 2010; Lee et al. 2006).

So far, there have been limited researches in Asian countries, including Malaysia, which explore the relationship between religiosity and youths’ risky behaviours, especially sexual activity. One Malaysian study showed that high religiosity is a protective factor for risky behaviours, such as alcohol, drug use and sexual activity (Noon et al. 2003). Another study among youth with delinquent behaviours found about 40 % of them had poor knowledge of their religion and low involvement with congregational religious activities (Ibrahim et al. 2012). A similar finding was observed in a case–control study in which personal religious practices, such as praying and reciting religious scriptures, were associated with lower involvement in cigarette smoking among schooling youths (Nor Afiah et al. 2012).

In trying to understand the relationship between religiosity and youth sexual activity, the challenge lies in choosing appropriate study instrument that is able to measure intended construct. Religiosity is a multidimensional construct that includes experience, belief (faith), knowledge and practice (Holdcroft 2006; Salleh 2012). Some researchers have limited religiosity to religious activity (extrinsic religiosity) and a general view of the importance of religion in one’s life (intrinsic religiosity) (Fatusi and Blum 2008; Hardy and Raffaelli 2003) or only intrinsic religiosity (Stulhofer et al. 2011). Some researchers have also included religious views on premarital sex, contraception, family religiousness and social religiosity in the religiosity questionnaire (Haglund and Fehring 2010; Gold et al. 2010; Ahmed et al. 2011; Burdette and Hill 2009; Edwards et al. 2008). However, till date, there has been no standard religiosity tool that examines the multidimensional construct of religiosity and youth sexual activity.

In this study, religiosity was measured in two dimensions: intrinsic religiosity and religious activity using the Duke University Religion Index (DUREL) questionnaire. Intrinsic religiosity refers to belief in the importance of religion in one’s life, while religious activity encompasses the frequency of performing private prayers and attending religious congregations. The DUREL was chosen because it is a short yet comprehensive five-item questionnaire and has been shown to be a valid and reliable tool to be used in epidemiological study in many countries including Malaysia (Koenig and Büssing 2010; Nurasikin et al. 2010).

As mentioned, the relationship between religion and youth sexual activity may be affected by gender, religious affiliations and settings. Given the fact that Malaysia is a conservative country, it was hypothesised that religion played a significant role on youth sexual activity and varied according to gender. However, it is not known how religion may have affect youth sexual activity and relying on one method may not be sufficient to understand this complex issue. By mixing quantitative and qualitative approaches, this study aimed to find out: what is the relationship between religiosity and youth sexual activity and how religiosity had influenced their decision on sexual activity.

Methodology

This study used an exploratory mixed methods research. The quantitative and qualitative studies were equally important and used to complement each other. Quantitative approach was used to determine the relationship between religiosity and youth sexual activity as it is the most cost-effective way to measure the prevalence of sexual activity and its factors including religiosity among college students. Concurrently, qualitative approach was used in order to have an in-depth understanding on how religion may have influenced sexual decision-making. Individualised in-depth interview would be the best approach as sexual activity and religiosity can be sensitive issues by some people. Both phases of the data collection took place between June and October 2013. Data were analysed accordingly, and the results were then converged to give a comprehensive understanding on the topic.

Sample Population and Study Location

This study involved students from 12 randomly selected colleges in Klang Valley, Malaysia. Klang Valley was chosen as the study site because it is where Kuala Lumpur, the capital city of Malaysia is located. In year 2013, there were 149 colleges centred in Klang Valley with students from all over the country.

Quantitative Approach

This was a cross-sectional survey. The urban colleges were randomly selected using computer-generated random number from a list of colleges registered under Ministry of Education of Malaysia. Administrators of each college were briefed about the study and the criteria for student selection. The inclusion criteria were students aged 18–22 years, unmarried, Malaysian and studying in colleges around Klang Valley. The exclusion criteria were students who were absent on day of data collection or those who did not state their sexual status in the questionnaire.

Due to logistical reasons and to minimise academic interference, the college administrator identified class that matched the study criteria. On the day of data collection, only the main researcher met the selected students in a prearranged classroom or hall. All students from each selected class were briefed on the study procedure and were requested to complete a set of self-administered questionnaire if they agreed to participate. Completed questionnaire without personal identification was sealed in an individual envelope and returned to the main researcher on the same day. Only the main researcher could get access to the completed questionnaires. Data collection was performed on different days at the various study sites. The sample size recruited for this study (1026) met the minimum sample required (948) to elicit differences across gender with confidence level of 95 %, power of study of 80 % and estimated prevalence of sexually active males of 10 % and females of 5 %.

The set of questionnaires contained questions that assessed sexual activity as the dependent variable and religiosity as the independent variable. Sexual activity was assessed using three questions: ‘Have you ever had sexual intercourse?’, ‘Over the past 12 months, have you had sexual intercourse with anyone?’ and ‘Are you actively having sexual intercourse with anyone?’, with ‘yes’ and ‘no’ responses. Answering ‘Yes’ to these questions indicated whether participants had ever been sexually active, had been sexually active over the last 12 months and were currently sexually active (up to the time of the survey).

Participants were asked to indicate their religious affiliation (Islam, Christian, Hindu, Buddhism, other faith or no religion). For the purpose of this study, the Duke Religion Index (DUREL) that captures intrinsic religiosity (i.e. the belief in the importance of religion in one’s life) and frequency of organisational and non-organisational religious activity was chosen for its brief yet comprehensive nature. Organisational religious activity was represented by the question ‘How often do you attend mosque, church, temple or other religious meetings?’, with responses ranging from never (1) to more than once a week (6). Non-organisational religious activity was represented by the question ‘How often do you spend time in private religious activities, such as prayer, meditation or reading the Quran or Bible?’, with responses ranging from never (1) to more than once a day (6). For the purpose of this study, these two questions were combined to denote religious activity, and thus, the total score for religious activity ranged from 2 to 12. Three questions measured intrinsic religiosity—‘In my life, I have experience the presence of the Divine or God’, ‘My religious beliefs are what really lie behind my whole approach to life’ and ‘I try hard to carry my religion into all other dealings in life’—with responses ranging from definitely not true (1) to definitely true of me (5). The total score for intrinsic religiosity ranged from 3 to 15.

The religiosity score was the summation score of the 5 items and ranged from 5 to 27. Higher scores indicated higher religious activity, intrinsic religiosity and total religiosity. In this study, the Cronbach’s alpha was 0.81 for the religiosity scale (5 items), 0.67 for the religious activity scale (2 items) and 0.88 for the intrinsic religiosity scale (3 items). The Pearson’s correlation between religious activity and intrinsic religiosity (r) was 0.54.

Data were analysed using Statistical Package for Social Science (SPSS) version 19.0 software. An independent t test was used to detect any differences in intrinsic religiosity and religious activity across gender. In subsequent analyses, religious affiliation was recategorised into Muslims and non-Muslims, which included Buddhists, Hindus, Christians and those with another religion or no religion. Multiple logistic regression analysis using the ‘Enter’ method was conducted to test the hypothesis that religious affiliation (non-Muslims as the referent group), religious activity and intrinsic religiosity influence sexual activity, while controlling for age and living situation (living with both parents as the referent group) for both genders. The level of significance was based on the adjusted odds ratio, with 95 % confidence intervals and p < 0.5.

Qualitative Approach

During the data collection of quantitative study, all students were invited to participate in the qualitative study and requested to write only their telephone number on a Qualitative Participant Form if interested to be interviewed. The form and the questionnaire set shared the same unique number to allow the researcher to select participants for the qualitative study according to their reported sexual status. To ensure maximum variation, the main researcher contacted students according to their sexual status, gender and religious affiliations. The main aim at this stage was to have almost equal numbers of sexually active and non-sexually active, male and female and Muslim and non-Muslim participants. In-depth interviews were arranged according to participants’ convenience.

Although there were 271 students interested to be interviewed, few problems were encountered such as some telephone numbers did not work, and some students changed their minds after completing the study in college or had started working. The recruitment process, interviews and data analysis took place, from June to October 2013. Recruitment of participants was discontinued upon reaching the point of data saturation. In total, 15 students underwent face-to-face interviews, either in a private room at their college, in a quiet café near their college or in the researcher’s car. All of the interviews were conducted by the main researcher using a semi-structured interview protocol. The interviews began with brief description about themselves and their family. The discussion of how religion had influenced their decision on sexual activity was triggered with the question: ‘In your opinion, what is your religion’s view on premarital sex?’ This was followed with a direct question on their sexual status and then to describe how religion may have affected their sexual decision. Member check or participant validation was done at the end of each interview, the researcher made a short summary of the interview, clarifying interpretations and inviting the interviewee to suggest any information, if necessary.

All the processes for qualitative data collection, from contacting students and arranging appointments for interviews to interviewing the students, were done by the same researcher. This allowed the establishment of rapport with students. They were more relaxed during the face-to-face interviews and more open in discussing their experiences of sexual activity. The interviews were recorded using a digital audio recorder, transcribed verbatim and imported into computer-assisted qualitative data analysis software, Atlas.ti (version 7), by the main researcher. The main researcher did the thematic analysis, and the emerging themes were compared across different sexual activity statuses. Apart from using the semi-structured interview protocol and member check at the end of each interview, several strategies were employed in order to improve the credibility of the findings. These were (1) repetitive process of listening to the digitally recorded data, transcribing word by word of the interviews and reread the transcribed data line by line multiple times, (2) double coding in which data were coded twice and the codes of the first and the second coding were compared, (3) external review of the generated themes by research team members. Any necessary amendments and clarification to the themes were made accordingly. The final themes were agreed upon by all members.

Ethics

The Research and Ethics Committee of the researchers’ institution reviewed and approved the research protocol, while the Malaysian Ministry of Education granted permission to sample students from registered colleges. Permission was also obtained from colleges’ administrators, and all participants were volunteers and gave informed consent. The original author of DUREL granted permission for the research team to use the questionnaire (Koenig and Büssing 2010). Anonymity was maintained throughout the research process. In reporting the qualitative findings, students were identified using numbers.

Results

Quantitative Results

Table 1 shows the socio-demographic profiles, sexual status and religiosity of the students. A total of 1026 students participated in the quantitative study, and more than half were female (57.7 %), Malay (62.8 %), Muslim (63.4 %) and living away from their parents (65.0 %). Regarding sexual status, 10.7 % reported ever having sex in the past, 8.7 % had been sexually active over the last 12 months, and 4.3 % were sexually active at the time of data collection. The mean age of the students was 19.7 (SD = 1.2 years), and the overall religiosity score of the students was high (19.5, SD = 4.8, range 5–27). The students scored high in intrinsic religiosity (12.6, SD 2.7, range 3–15) and low in religious activity (7.0, SD 2.7, range 2–12).

Table 1 Socio-demographic profiles, sexual status and religiosity score of the participating students (n = 1026)
Full size table

Male students showed significantly higher total religiosity scores (t = −3.80, p < 0.001) and religious activity (t = −6.1, p < 0.001) compared to female students (Table 2). However, there was no significant difference in intrinsic religiosity among male and female students. Based on these findings, subsequent analysis on the relationship between religiosity and sexual activity was carried out according to gender.

Table 2 Difference in DUREL religiosity score by gender
Full size table

Table 3 shows the multiple regression analyses that were conducted to determine the relationship of religion and religiosity with sexual activity in male and female students, while controlling for age and living situation. In male students, religious affiliation, religious activity and intrinsic religiosity were not significant in predicting all the three sexual statuses. In female students, the only significant relationship was between religious activity and current sexual activity at the time of the survey. The logistic regression model was statistically significant, χ 2 (5) = 11.34, p < 0.05 and explained (Nagelkerke R 2) 10.0 % of the variance in current sexual activity. Engaging in religious activity reduced the odds of sexual activity among the female students (adjusted OR = 0.67, CI = 0.47, 0.95, p = 0.02).

Table 3 Multiple logistic regression analysis of different sexual status by gender
Full size table

Qualitative Results

Of the 15 students interviewed, 7 were sexually active and 8 were not sexually active. Their ages ranged from 18 to 22, 9 were female, and 6 were male. There were 10 Muslims, 3 Hindus and 2 Buddhists participating. Thirteen students were living away from their parents, and only two were living with both parents.

Two main themes to explain how religion had affected their sexual decision emerged from the thematic analysis: (1) having faith in religion prevents risky sexual activity. (2) Strong sexual desires and ignoring religion perpetuate sexual activity.

Having Faith in Religion Prevents Risky Sexual Activity

Among non-sexually active students, both males and females believed having faith in religion had prevented them from engaging in risky activities, including having premarital sex. For example, N1, an 18-year-old Muslim female mentioned: ‘I think having strong faith in religion is important…. If the person is religious, he or she won’t do that [have sex]’. She justified her decision to not have an intimate relationship at her age as it is not allowed. She believed she had to be good by not doing prohibited activities in order to have a good husband and family. She commented:

I am not ready for marriage, so why should I have a boyfriend? … From a religious perspective, this is not right. … I want to have a good husband, and Allah has promised that a good man is for a good woman. So if I want to have good children, but if I am not good [do prohibited activities], how can I get a good husband? … [I] need to think about sin, too. (N1, an 18-year-old Muslim female, not sexually active)

N15, a 21-year-old Hindu believed that virginity is from God, and it needs to be kept until marriage.

N15::

Girls need to keep their virginity. For me, virginity is love from God

Interviewer::

For whom do you want to keep the love?

N15::

For my husband, when I get married

Religion gave participants a sense of control, and some used religion as a protection against deviant behaviours, including risky sexual activities as narrated below.

When we perform our prayers regularly, it can control us, so we don’t think about other things. If a person doesn’t pray, he will think of many things, including sex. (N11, a 19-year-old Hindu male, not sexually active)

A similar view was mentioned by N4, a 20-year-old Muslim male. He believed that his obligatory five-times-a-day prayers, together with additional religious activities, were able to guide and control him from even thinking about having sex.

Interviewer::

What do you think about those young men who have premarital sex?

N4::

They shouldn’t, but I don’t know how they even think about doing it [sex]. My ustaz [religious teacher] said it is because they have dark hearts

Interviewer::

So how do you look after your heart and don’t think about sex?

N4::

I really pay attention to my solat [prayers]. I do extra solats [prayers], too. Before prayers, I will recite the zikr [a prayer] and read the Quran

Prayers and religious activities, such reading the Quran or other religious scriptures and attending mosque, temple or religious discussions, were regular activities for many of the non-sexually active students. Whenever they were at home, the activities became part of family activities, and parents played an important role in monitoring students’ religious activities. One participant described:

When it’s time for prayer, my mum would remind us to get ready and do our prayer. My dad, too. Sometimes, he brought the whole family to mosque. We will be scolded if we don’t pray.” (N4, 20-year-old Muslim male, not sexually active)

In contrast, none of the sexually active students, whether Muslim or Buddhist, faithfully practised religious activities. Some only performed the prayers when they were at home with their families, while others rarely performed obligatory prayers. The following are example of narratives regarding students’ religious activity.

I go to temple when I am at home and not here [at college]. (N13, 22-year-old Buddhist female, sexually active)

My mother will get angry if I don’t pray. If at home, I have to pray five times [a day]. When at college, I don’t do [that]. Sometimes, I don’t pray the whole day or after my class. I just don’t do [prayer]. (N10, 19-year-old Muslim female, sexually active).

Some of the Muslims believed that their loose connection with God (Allah) had made them go astray. Some mention:

It’s because of my weak faith. My religious activities are little. My prayers are not enough or perhaps incomplete, so God doesn’t see them, making me to do things, following the Devil. (N2, 20-year-old Muslim male, sexually active)

I skipped my solat [prayers] …. [If my parents asked], I lied. … [The] solat was not able to control me. (N7, 18-year-old Muslim male, sexually active)

Strong Sexual Desires and Ignoring Religion Perpetuate Sexual Activity

All Muslim and Hindu students knew that premarital sex is forbidden in their religions. However, two Buddhist students were unsure about their religion’s view of premarital sex. Despite knowing the religious prohibition, the sexually active students felt strong sexual desire whenever they were with their sexual partners, which drove them to have sexual intercourse. Some of excerpts of their statements are as follows:

It is forbidden in Islam to have sex before marriage. … But when I have sex, I don’t think about anything [religion]. (N8, 19-year-old Muslim female, sexually active)

I know it is sinful, but when you are with your partner…there’s the third person, the evil [the sexual desire]. (N10, 19-year-old Muslim female, sexually active)

Sexual arousal was more influential than their belief in religion, and they were willing to accept the sin. N7, an 18-year-old Muslim male, mentioned: ‘I just cannot control my sexual desire. I got addicted. I just accept [the sins]’. A Muslim male admitted usually feeling guilty after his sexual activity, but that did not deter him from continuing the sinful activities, to which he attributed his strong sexual desire and the need to satisfy his sexual needs (having pleasure).

I trust Allah and sins, but when we do the thing [sex], we just want to enjoy it and satisfy the sexual desire. We don’t think anything else [about sin]. Once it’s over, only then we have regret. (N2, 20-year-old Muslim, male, sexually active)

In contrast, all the non-sexually active youth denied having the intention or desire to have sex and asserted that their faith in religion guarded their activities. N4 avoided anything that might lead to the arousal of sexual desire, including viewing pornography. N1 chose not to have any intimate relationship, and N6 would demand marriage if she received a sexual request as both believed that premarital sex is wrong in their religion.

I don’t want to try…. I know it is a sin, and it is not a minor sin…. I don’t even want to watch pornography. I went to religious school, so I know it is wrong and sinful, so I do not want to do it. (N4, 20-year-old Muslim male, not sexually active)

Why do I need a boyfriend? My parents give me enough love and care? Why should I find it from someone else? From a religious perspective, it is not right, anyway. (N1, 18-year-old Muslim female, not sexually active)

I know it is prohibited and sinful in our religion. Even if we really like the person, then get married first. (N6, 21-year-old Muslim female, not sexually active)

Reconciliation of the Findings

In the questionnaire, the students had high scores for intrinsic religiosity (faith) and low scores for religious activity. Religious activity was found to be a protective factor for current sexual activity (at the time of the data collection) among female students but not male students. In the interviews, having faith in religion and engaging in religious activities portrayed to be protective for non-sexually but not sexually active students. All of the sexually active students admitted to not faithfully practicing their religious activities. Although all of them knew their religion’s prohibition on premarital sex, they tended to disregard their religion’s teaching because of strong sexual urges, desires and indulged in premarital sexual activity.

Discussion

In general, the youth in this study showed high intrinsic religiosity but had low religious activity scores. A similar finding was noted among youth from the USA. Many of them believed in God or a universal spirit and admitted the importance of religion to them, but not all of them were regularly involved in religious activity (Rostosky et al. 2004). Although traditionally, religious affiliation, such as being Muslim or Hindu, has been associated with less involvement in deviant sexual activities (Adamczyk and Hayes 2012), our study failed to support this finding. Regardless of youths’ religious affiliation, religiosity played little role in their sexual activity. This could be because of the effects of modernisation, which have led to the erosion of traditional, conservative values. The younger generation is more liberal, and religion has become less important in their life. This pattern was observed among Australian youth, for whom religion had little importance in determining their sexual activity (Siebold 2011). A study of 126 youths in London ages 15–18 with different religions showed that religion might guide their lives, but factors, such as gender norms, peer pressure and sexual needs, might override their religious beliefs and influence their sexual activity (Sinha et al. 2007a, b).

Similar to other studies conducted elsewhere, this research found a gender-specific effect in the role of religion in youth sexual activity. A higher protective influence from religion and, in particular, religious activity was noted among female youths compared to male youths. Religiosity was a protective factor for sexual activity among females in studies done in the USA (Edwards et al. 2008; Gold et al. 2010) and Croatia (Stulhofer et al. 2011). Religious or moral values acted as the motivating factor in female youths who adopted non-permissive attitudes towards premarital sex (Rostosky et al. 2004; Ahrold et al. 2011). One of the possible reasons why males are less affected by religiosity is because there is greater emphasis in almost all religions on females keeping their virginity until marriage (Ahrold et al. 2011). This was observed in one of the female youths in our study, N15, who valued virginity as something precious from God.

Strong sexual urges and desires had made many of the sexually active youth ignore their religious teachings. They were impulsive with their sexual desire and continued to practise sex, especially whenever they were with their sexual partners. Sexual desire or lust was found to be one of the reasons why youth maintained premarital sexual activity (Subhi et al. 2012). Once a person is controlled by desire, it is very likely the person will not behave in accordance with religious teaching (Ahmad et al. 2007). One of the ways in which religion can be effective in assisting youth to adopt a healthier, safer lifestyle is by encouraging them to get involved in religious activities.

As in the quantitative study, all the youths were noted to have low rates of involvement in religious activities. Frequent religious activity and having conservative attitudes in accordance with one’s religion were associated with lower chances of risky sexual behaviour (Haglund and Fehring 2010). Those with high religiosity and valuation of religion in their life tended to delay their sexual debut (Hardy and Raffaelli 2003; Burdette and Hill 2009). This was also observed in interviews with the non-sexually active youth, who reported that their strong faith and commitment to their religion through praying and reading religious scriptures had prevented them from indulging in risky sexual activity.

Those who understand their religion adhered to religious teachings and behaviours in any situation (Ismail et al. 2011). Religious involvement increased self-esteem and self-efficacy. It provided a sense of personal control against deviant behaviour while reducing the likelihood of engaging in premarital sexual activity and other risky behaviour. As mentioned by the non-sexually active youths in the qualitative study, having a strong faith in religion, knowing what is wrong in the religion and using religion as a means to control their activities had prevented them from practising premarital sexual activity. Strong intrinsic religiosity put them in control of their behaviour. They saw the long-term impact and risks of current illicit activities, rather than only focusing on the short-term pleasures (Klanjšek et al. 2012). Knowing they were being monitored by their God reduced their risk of involvement in deviant behaviour (Klanjšek et al. 2012).

As this was a cross-sectional study, it lacks data on temporal relationships, and students’ religiosity when they first engaged in sexual activity is not known. Their religiosity level, which could be high or low, might have significantly affected their first involvement in sexual activity and subsequently changed over time due to other factors. However, this dynamic could not be ascertained in this cross-sectional study. A future longitudinal study with multiple measurements of religiosity levels from the sexually naïve stage to sexually activity could capture changes in the level of religiosity during this process. The involvement of college administrators in selecting group of students for data collection was unavoidable. Nevertheless, confidentiality of the students was protected as college administrators were not involved during data collection and only the main researcher could get access to the sealed responded questionnaires.

Conclusion

This study highlights the importance of adhering to religious activities in preventing female youth from continuing their sexual activity. However, religion was not able to restrain the sexual activity among male youth. Once youth has experienced sexual activity and when sexual urges and desires were beyond control, religiosity might not work and become a less important protective factor. Interestingly, faith in religion emerged as an important theme that prevented sexual activity among sexually naïve youth. A future randomised control trial that look into the effectiveness of strengthening faith in religion to prevent premarital sex may be carried out in order to strengthen this finding and making it applicable in primary preventive programme.

As for the secondary prevention, the intervention programmes need to be gender specific especially when using religious approach. Focussing on religious activity may be beneficial for sexually active female youth. However, different approach needs to be used with sexually active male youth. In both male and female sexually active youth, they need to be empowered and taught how to manage their sexual desire in a healthy and positive way.