Sociology and Social Informatics

Размещено на http://www.allbest.ru/

Sociology and Social Informatics



Introduction

Despite the proven effectiveness of sex education (Pearlman et al. 2002; Beshers 2007; Jennings et al. 2014; Shulgina 2015), there is still no compulsory program which would concern issues of sexual development in Russia. Russian state guarantees the implementation of measures to prevent the dissemination of human immunodeficiency viruses (HIV) and sexually transmitted diseases (STDs), however, most of these activities are one-time and cannot be considered a full course of adolescents' sexual education. In Russia, there are a few studies on the sexual health of adolescents and the assessment of adolescents' sexual knowledge about contraception and prevention (Zhuravleva 2004; Krivoruchko and Lipskikh 2013; Petrova 2014; Soldatova and Rasskazova, 2014), it remains unknown what level of sexual literacy teenagers have, and which false knowledge and perceptions teens have. Thus, it is important not only to evaluate the level of sexual health literacy of adolescents but also to assess the prevalence and effect of false knowledge in order to understand how correct and false knowledge is associated with the sexual behaviour of adolescents. Sexual literacy is a broad concept that includes many aspects. In this study, sexual health literacy is studied in terms of knowledge about contraceptive methods and their effectiveness in preventing unwanted pregnancy and STDs.

The purpose of this study is to describe how sexual literacy is associated with self-efficacy of condom use and safe sexual behaviour among adolescents from Saint-Petersburg's vocational schools. The research questions are the following:

1. How the high level of sexual knowledge is related to self-efficacy of condom use and safe sexual behaviour (the use of contraception)?

2. How the presence of false knowledge is associated with self-efficacy of condom use and safe sexual behaviour (the use of contraception)?

The study sets the following tasks:

? to analyse how sexual literacy is related to self-efficacy of condom use;

? to analyse how sexual literacy is related to safe sexual behaviour;

? to analyse how the presence of false knowledge is related to self-efficacy of condom use;

? to analyse how the presence of false knowledge is related to safe sexual behaviour.

The final thesis consists of 3 chapters. The first chapter describes the existing research dedicated to the predictors of safe and risky sexual behaviour. A separate block in this chapter discusses the work of Russian scientists who studied the sexual behaviour of adolescents in Russia. The Health Belief Model is also described, the elements of which I use as an explanatory mechanism in my work.

The second chapter describes the research methodology and data processing. The data includes 1001 students aged 17 to 21 from 13 Saint-Petersburg's vocational schools. All data was collected by the Scientific and Educational Laboratory Sociology of Education and Science (Higher School of Economics, St. Petersburg). I myself took part in collecting data and conducting surveys. During the data processing, several new variables were elaborated (mother's education, sexual health literacy, false knowledge, perceived severity (`Shame' and `Terrible consequences'), risk perception, sexual experience, perceived susceptibility, sexual behaviour, self-efficacy of condom use and the variable that shows the amount of people with STD that an adolescent is familiar with). To analyze the obtained data, two models were created. For both models, control variables of age, gender, and maternal education were accounted for. For the first model, the dependent variable is self-efficacy of condom use. The following variables were used as independent variables: sexual experience, sexual health literacy, false knowledge, risk perception, perceived severity (`Shame' and `Terrible consequences'), perceived susceptibility, knowing someone with STDs. To build the model, a linear regression was applied. For the second model, the dependent variable is sexual behaviour (the use of contraception during the last sexual intercourse). The following variables were used as independent variables: sexual health literacy, false knowledge, risk perception, perceived severity (`Shame' and `Terrible consequences'), perceived susceptibility, knowing someone with STDs. To build the model, binomial linear regression was used, the odds ratio with 95% confidence intervals was calculated.

The third chapter describes the results of the models. The conducted research enables us to make several conclusions. First, according to the first model, the higher the risk assessment is, the higher the probability that a teenager's self-efficacy of using a condom gets. Second, the higher the level of sexual literacy a teenager has, the greater the probability that his/her self-efficacy of condom use is. Moreover, females tend to rate their personal risks higher than male adolescents do and female students also evaluate their personal self-efficacy higher. However, the presence of false knowledge about different types of contraception and their effectiveness did not show a significant association with self-efficacy. The second model gave the following results. First, if a male appreciates the risk of an unplanned pregnancy in the case of sexual intercourse without using a condom, then he has better odds of using contraception during sexual contact. What is more, the higher a teenager's self-efficacy of condom use is, the higher the chance that he/she uses contraception gets. Finally, sexual health literacy and false knowledge did not produce significant effect on sexual behaviour.

1. Literature Review

virus sexual safe

The concept of sexual literacy includes a wide range of knowledge. In 1995 The National Commission on Adolescent Sexual Health identified several important components that constitute sexual literacy: the ability to create and maintain interpersonal relationships, the ability to value one's body, the ability to interact with people of both sexes respectfully, and the ability to express love and closeness in ways which correspond to one's values ?(Haffner 1995). The concept of sexual literacy was later expanded (Satcher 2001) to include the ability to evaluate, understand and weigh the risks, responsibilities, and consequences of sexual activities. Moreover, it includes freedom from sexual violence and discrimination, as well as the ability to enjoy sex life (Tolman, Striepe, and Harmon 2003).

As for sexual risk health literacy, it includes not only knowledge and behaviour but also reflects how motivation and the ability to understand and evaluate information are used to solve problems with sexual risk behaviour (Vongxay et al. 2019). The quality and accessibility of sexual and reproductive health information (Christiansen, Gibbs, and Chandra-Mouli 2013; Chandra-Mouli et al. 2014) have a major impact on adolescents' ability to understand and apply knowledge of sexual literacy.

In research, sexual health literacy is seen as knowledge on HIV/AIDS and its transmission (Kayiki and Forste 2011), knowledge of the types and transmission of STDs (Langille et al 1998), proper condom use (Langille et al 1998, Kayiki and Forste 2011), and knowledge of how to prevent an unwanted pregnancy (Bankole et al. 2007). Thus, sexual literacy is a broad concept that can be studied from several perspectives.



1.1 Predictors for safe and risky sexual behaviour

Today, it remains impossible to speak accurately about the relationship between sexual literacy of adolescents and changes in their sexual behaviour (Kotchick et al. 2001). Nevertheless, several studies emphasize that adolescent knowledge is positively associated with condom use (Mahat and Scoloveno 2010; Kayiki and Forste, 2011; Swenson et al. 2010). In the study conducted by Thanavanh et al. (2013), adolescents who showed a medium and high level of knowledge were significantly more likely to practice safe sexual practices. Thus, adolescents with higher knowledge later start their sex lives, are more aware of HIV and are more likely to use condoms (Berten and Van Rossem 2009). Research by Mahat and Scoloveno (2010) and Tapia-Aguirre et al. (2004) showed that the more adolescents knew about HIV/AIDS, the more likely they were to use condoms. Adolescents who show a high level of sexual literacy not only use condoms but also understand the benefits that they get from the use of contraception (Terry et al. 2006). Moreover, increased knowledge on HIV is also associated with a higher likelihood of a teenager being tested for HIV (Swenson et al. 2010). In another study (Champion, Harlin, and Collins 2013) that examined girls and boys who went to the clinic for treatment for STDs, the data showed that those who had heard about STDs received more treatment (36%) than those who had not heard about STDs before infection ( 26%).

Researchers note that even though adolescents may know sexual literacy, this does not mean that this knowledge is complex enough (Bankole et al. 2007). Bankole et al. (2007) emphasize that despite being highly aware of HIV, adolescents lack in-depth knowledge about HIV prevention and transmission. The same goes for condom use: if adolescents have good enough knowledge about condom use, they are more likely to use it (Kayiki and Forste 2011). Thus, more accurate condom knowledge (Kayiki and Forste, 2011), such as a demonstration of condom use (Bankole et al. 2007), has the greatest impact on the correct use of condoms by adolescents.

Despite all of the above, the relationship between the knowledge about sex and sexual health and sexual risk behaviour is rather complex (Kotchick et al. 2001). More precisely, the gained knowledge and facts can increase the level of knowledge among adolescents, but this does not always lead to a change in their behaviour (Levinson 1995). Lou and Chen (2009) have shown that sexual health knowledge does not always lead to a more frequent use of condoms and other safe sex behaviour practices. As other researchers explain, knowledge itself may not be sufficient to form safe sexual behaviour - intentions and perceptions of social norms are equally important (Langille et al. 1998). Another explanatory mechanism between knowledge and risk behaviour may be the level of a teenager's moral reasoning, which can reduce or increase his/her sexual risk behaviour (Hubbs-Tait and Garmon 1995). Rock (2005) explains the inaccuracy of the relationship between knowledge and sexual behaviour by the fact that the existing research focuses only on objective knowledge. If one accounts not only for objective but also perceived knowledge, it can effectively help change the sexual behaviour of adolescents and increase their use of condoms (Rock 2005).

Even adolescents who have a high level of knowledge may exploit false beliefs about safe sexual behaviour (Tavoosi et al. 2004). Tebourski and Benalaya (2004) and Huang et al. (2005) point out that adolescents tend to misjudge HIV transmission and have false beliefs about condom use. For example, in studies by Tavoosi et al. (2004), Thanavanh et al. (2013), Tung, Ding, and Farmer (2008), respondents noted that HIV can be transmitted through mosquito bites, from shaking hands, toilet seats, and public swimming pools. In a study by Bhattacharya, Cleland, and Holland (2000), adolescents also believed that only a male could become infected with HIV. Such misconceptions can lead to underestimation of risk, as well as hinder timely access to medical care (Bhattacharya, Cleland, and Holland 2000). With regard to contraception, misconceptions about condom use are more common among older teens who have already had sexual experience (Crosby and Yarber 2001).

The awareness of risk and perception of personal vulnerability can lead to protective behaviour: condom use can be affected by a case of illness or death from AIDS of someone familiar to a person (Ekanem et al. 2005). In a Kayiki and Forste (2011) study, adolescents who was acquainted withsomeone infected with AIDS had a 113% higher chance of using condoms. Moreover, people who interacted with someone who had HIV/AID or someone who would had died because of it, were more likely to be tested on HIV (Opt and Loffredo 2004).

Adolescents who have no sexual experience are more optimistic about avoiding an unwanted pregnancy which can be an element of risky behaviour (Whaley 2000). Moreover, Sutton et al. (2011) emphasize that there is a mismatch between risk perception and adolescent sexual behaviour. Chapin (2001) and Sutton et al. (2011) attribute this to an optimistic bias: although adolescents engage in sexual risk behaviour, they underestimate their vulnerability to the negative consequences of risky sexual behaviour.

An equally important predictor of safe sexual behaviour is self-efficacy - a person's confidence that he/she will be able to take preventative measures, for example, confidence in his/her ability to use a condom (Wiener, Battles, and Wood 2007). A study by Slonim-Nevo and Mukuka (2005) suggests that low self-efficacy about AIDS prevention is largely associated with involvement in risky sexual behaviour. Self-efficacy also applies to condom use (Wiener, Battles, and Wood 2007). The teenager's confidence that he/she can use condoms on their own is one of the most important factors in preventing STDs (Wiener, Battles, and Wood 2007).

1.2 Sexual literacy of adolescents in Russia

In Russia, there are very few studies that would examine the relationship between adolescent knowledge about sex and sexual health and risky sexual behaviour. Nevertheless, researchers note that sexual health continues to deteriorate: according to Petrova (2014), the number of 15-17 years adolescents infected with syphilis have increased by almost 25%, and gonorrhoea by 57%. Petrova (2014) emphasizes that if the situation does not change, then many couples in the future will not be able to have children. As for contraception, about 15% of teenage girls are not protected, and 16% are not at all familiar with various methods of contraception (Petrova 2014). Regarding concerns about STDs, only 34.5% of those surveyed said they had concerns during sex, and rejection of sex due to fear of contracting STDs was indicated by only 9% of guys and 1% of girls (Rogacheva, Malikova, and Zakharov 2015). Researchers note that such lowered fears may be associated with a high level of trust: 75% of respondents noted a high degree of reliance between them and their partner (Rogacheva, Malikova, and Zakharov 2015). Despite the fact that adolescents are assumed to know that they need to be tested every year for sexually transmitted diseases, only 28.9% of adolescents who have sex are likely to take a test (Filonova 2012). All in all,Russian researchers focus more on measuring the knowledge of adolescents and describing their sexual behaviour separately, They emphasize that most adolescents have extremely low knowledge and are prone to risky sexual behaviour (Zhuravleva 2004; Rogacheva, Malikova, and Zakharov 2015), but the existing studies dedicated to Russian teens do not explore the relationship between the two interconnected factors.

1.3 Theoretical framework: Health Belief Model

The Health Belief Model (HBM) is a cognitive model for predicting behaviour and changes in human behaviour (Rosenstock, Strecher, and Becker 1988, 1994). Health Belief Model is used to understand risk behaviours for health (Conner and Norman 2007), including risky sexual behaviour (Brown, DiClemente, and Reynolds 1991). HBM suggests that preventive types of health behaviour depend on several factors (Rosenstock, Strecher, and Becker 1988, 1994):

? how people assess their perceived susceptibility to adverse health outcomes (perceived susceptibility);

? how seriously do they perceive the adverse effects on their health and all related outcomes (perceived severity);

? how people perceive the benefits of preventative behaviour (perceived benefits);

? what are the alleged barriers to preventive behaviour (perceived barriers);

? how people perceive self-efficacy in implementing preventive behaviour (self-efficacy).

Thus, the general idea of theHealth Belief Model is the following: for people to take some kind of action to prevent and control their health, they must consider themselves susceptible to this disease (Rosenstock, Strecher, and Becker 1988, 1994). The probability of applying protective measures is also influenced by whether a person considers himself/herself capable of performing protective behaviour and how seriously he/she assesses the health consequences (Rosenstock, Strecher, and Becker 1988, 1994). Moreover, one must understand that the benefits of protective behaviour outweigh the barriers that impede its implementation (Rosenstock, Strecher, and Becker 1988, 1994).

The Health Belief Model is used as an explanatory mechanism for sexual risk behaviour in many studies. Studies find support for several components of HBM (Brown, DiClemente, and Reynolds 1991; Zak-Place and Stern 2004). A study by Roye and Hudson (2003) showed that teenagers tend to take preventive measures during sex when they understand how STDs can affect their lives. Concerning self-efficacy, Strecher, Champion, Rosenstock (1997) also indicate that self-efficacy in condom use was positively associated with a lifelong condom use. A study by Winfield and Whaley (2002) emphasizes that perceived barriers to condom use are significant predictors of condom use. However, it is important to consider that though condom use can be attributed to several components of Health Belief Model, other factors may also influence condom use (Winfield and Whaley 2002).

The aim of the study is to examine how sexual literacy is associated with safe sexual behaviour and self-efficacy of condom use. One of the most important principles of Health Belief Model is the concept of susceptibility to the disease (a person must believe that he/she is at risk and can get sick). Even if a teenager has relatively high knowledge about various types of contraception, if he/she does not feel that he/she is at risk of STD or unwanted pregnancy, then his/her perception of risks is much less, which can lead to his/her risky sexual behaviour. As it is known from the literature, a different situation occurs when an adolescent communicate with someone who has STD or would have died from AID, because this experience can increase the awareness due to a sense of personal vulnerability (Ekanem et al. 2005; Kayiki and Forste 2011). In case if a teenager possesses false knowledge, in this case, he/she is more likely to not feel at risk and will not be able to seriously assess the health consequences, which also leads to risky sexual behaviour. Moreover, due to the low confidence that he/she will be able to use contraceptives (self-efficacy), even with high knowledge, a teenager can still be involved in risky sexual behaviour. Thus, I will use perceived susceptibility, perceived severity, and self-efficacy as explanatory elements of HBM.



2. Methods

2.1 Data

The sample of this study includes adolescents from 13 Saint-Petersburg's vocational schools aged 17 to 21 (average age 18 years, 37% girls). It was chosen to use quantitative methods, so a survey was completed, with the information about teenagers' knowledge and sexual behaviour gathered during a longitudinal research of sexual behaviour of students. All data was collected by the Scientific and Educational Laboratory Sociology of Education and Science (Higher School of Economics, St. Petersburg). The research project was evaluated by the HSE Committee on Interuniversity Surveys and Ethical Assess of Empirical Research as conforming to ethical standards. Written consent was obtained from students, their parents, and college directors to participate in the study. I participated in data collection and conducted student surveys at colleges in St. Petersburg

Secondary data included 1052 students. Students who did not answer most of the questions were removed. At the end of the questionnaire, the teenager had to indicate how honest he/she answered the questions. The question “How many questions of the questionnaire you answered honestly” have 4 categories of answers: (1) `All questions', (2) `Most of the questions', (3) `Only some', (4) `Not one'. If most of the questions were answered unfairly (based on the personal assessment of the teenager), this student was removed. Also, only students from 17 to 21 years old remained in the final sample, since they fall into the studied category of adolescents (less than 17 years old students were not in the database initially). Thus, the final data used in the research amounted to 1001 students.



2.2 Measures

Sexual Health Literacy and False Knowledge

The questionnaire handed to the respondents listed methods of contraception (interrupted intercourse, condoms, calendar method, chemical contraception, hormonal birth control pills, hormonal contraceptive patch), and a respondent was suggested to respond whether each type of contraception protects from STDs and unwanted pregnancy. The student could choose between (1) `Yes', (2) `No', or (3) `I do not know'. Thus, an expected sexual health literacy was calculated, which reflects adolescents' knowledge (the number of correct answers to the posed questions) about the effectiveness of contraceptive methods in preventing STDs and unplanned pregnancies. The Sexual health literacy summarizes the correct responses of adolescents regarding knowledge of the effectiveness of contraceptive methods in preventing STDs and unplanned pregnancies. To make the survey's results more complex, the question “Who do you thinkcan get STDs?” was added to the Sexual health literacy test. A teenager could answer: (1) `Only girls', (2) `Only boys', (3) `Girls and boys', or (4) `No one'. In the case when a teenager noted that (3) “Girls and boys” can get STDs, this answer added a point to his/her sexual health literacy rate.

Sexual health literacy rate can take values from 0 to 13, where 0 is the minimum of correct answers, and 13 is the maximum of correct answers given by a student.

False knowledge was calculated as a sum of incorrect answers on the same questions. The measure of false knowledge, thus, reflects the false knowledge of adolescents (the incorrect answers to the posed questions) about the effectiveness of contraceptive methods in preventing STDs and unplanned pregnancies. What is more, if a teenager answered that (1) `Only girls', (2) `Only boys', or (4) `No one' can get STDs, this answer added a point to his/her false knowledge rate.

False knowledge can take values from 0 to 13, where 0 is the minimum of incorrect answers, and 13 is the maximum of incorrect answers that a student could have. The scales for false knowledge and sexual literacy are not the same. False knowledge and sexual health literacy show a weak correlation (App. 1, 0.06*).

On the questions about contraception, a teenager could also answer that he/she does not know anything about the listed methods and its effectiveness in protecting from STDs or unwanted pregnancy or both, and in this case such an answer was not equated to either sexual health literacy or false knowledge rate. For the answers of the category `I don't know', the variable `Lack of knowledge' was created. However, this variable shows a strong correlation with Sexual health literacy (App. 1, -0.84**), so it was removed from further analysis.

Sexual experience

When filling out the survey, teenagers noted whether they already had sex. This question had only 2 possible answers: (1) `Yes', (2) `No'. If a teenager noted that he/she already had sexual experience, it was encoded as 1. If a teenager noted that he/she had not yet had sex, it was encoded as 0.

Sexual behaviour

Sexual behaviour reflects whether a teenager used any of the contraceptive methods that protect against unwanted pregnancy during the last sexual intercourse. A teenager should have indicated, who exactly took the responsibility for the use of contraceptives: (1) `I took', (2) `My partner', (3) `We both', (4) `We did not use any contraception'. In the case when either a (1) teenager, (2) his/her partner, or (3) both took responsibility for the use of any type of contraception, it was encoded as 1, which means that the teenager was using contraception against unwanted pregnancy during his/her last sex. In the case of choosing the answer (4) `We did not use any contraception', it was coded as 0, which means that the teenager did not use any contraception against unwanted pregnancy during the his/her sexual contact.

Self-efficacy of condom use

Self-efficacy of a condom use reflects the perception of self-efficacy regarding the use of condom in three situations. First, a teenager answered how confident he/she is that he/she can stop during sex in order to put on a condom. Second, a respondent noted how confident he/she is that he/she could take care of having a condom in advance (prior to a sexual intercourse), Finally, students marked the level of their belief in how confident they are in being able to refuse having sex without a condom. These questions had 5 categories of possible answers: (1) `Not at all sure', (2) `Not very sure', (3) `50 to 50', (4) `More or less sure', (5) `Absolutely sure'. There was also a response category (6) `I never use a condom', but since it does not reflect self-efficacy of a condom use, in the case when a teenager answered one of 3 questions (6) “I never use a condom” the answer was not considered further. Self-efficacy of a condom use rate was calculated as the average value of answers to the three questions.

Risk Perception

Risk perception calculates the average value of two questions. In the first question a teenager should have noted what the probability of becoming pregnant/of partner's pregnancy in case of one-time unprotected sex is, according to his/her mind. In the second question teens were suggested to estimate the probability of becoming pregnant/of partner's pregnancy in case of unprotected sex for a whole month. The two questions had 5 similar potential answers: (1) `Almost no chance', (2) `There is a chance, but small', (3) `50 to 50', (4) `Quite possible', (5) `Almost certainly'.

Perceived Susceptibility

The perceived susceptibility variable reflects the answer to the question of how a teenager evaluates the probability that he/she will get any of sexually transmitted diseases. The provided options of answers were the following: (1) `Very high', (2) `Pretty high', (3) `Pretty low', (4) `Almost none at all'. Answers (1) `Very high', (2) `Pretty high' were combined together and coded as 1, answers (3) `Pretty low', and (4) `Almost none at all' were combined together and coded as 2.

Perceived Severity

In order to identify one's perceived severity, the participants of the study were asked what would happen in case of an unplanned pregnancy regarding them or their partner. The perceived severity was divided into two variables. The first variable of perceived severity called `Shame' and it summarizes the average score of answers to the two following statements: `If I get pregnant/my girlfriend gets pregnant, it will be a shame for me' and `If I get pregnant/my girlfriend gets pregnant, it will be a shame for my family'. Both assertions had 5 categories of answers: (1) `I don't agree', (2) `I rather disagree', (3) `I don't agree nor disagree', (4) `I rather agree', (5) `I agree'. Thus, variable `Shame' is an average score of student's answers on these two statements.

The second variable of perceived severity called `Terrible consequences' and it sums up the average for the two other statements: `If I get pregnant/my girlfriend gets pregnant in the near future, it will be just awful' and `If I get pregnant, I will make an abortion/If my girlfriend gets pregnant, I will persuade her to have an abortion'. These questions had the same 5 possible choices: (1) `I don't agree', (2) `I rather disagree', (3) `I don't agree nor disagree, (4) `I rather agree', (5) `I agree'. Thus, variable `Terrible consequences' is an average score of student's answers on these two questions.

Knowing someone with STDs

The respondents were asked how many people with STD do they know. This was an open question, where the teenager was suggested to write the number of people with STDs familiar to him/her. Most responded that they did not know a single person or were familiar with 1-2 people with STDs. Therefore, this variable was converted from numeric to categorical one with two levels: if a student responded that he/she does not know a person with STDs, it was coded as 0, and if a participant marked that he/she knows one or more people with STDs, his/her answer was coded as 1.

Social-demographic characteristics

Age, gender (as a biological sex of the student: 0 - female, 1 - male) and a participant's mother's education were gathered during the survey. The question of mother's education reflects the presence of higher education: 1 for the cases when a respondent's mother possesses a higher education, and 0 when she lacks higher education, or she possess an incomplete higher or lower higher education, or a respondent dos not know his/her parent, or the mother's education is unknown.

2.3 Data analysis

In order to analyze the obtained data, two models were elaborated.

The first model uses self-efficacy as a dependent variable and, thus, reflects the adolescents' self-efficacy regarding the use of condoms. To test this model, answers of all 1001 students are used, regardless of whether the respondents to the survey had sex or not.

To build the model, a linear regression was applied. For this, packages from R Studio were used. The distribution of the self-efficacy of condom use is not linear, respectively, this can give a biased estimate of precision, and the coverage coefficient of confidence intervals may not correspond to their real level. To resolve this issue and present a more coherent result, a robust estimator of variance was utilized. Estimates were calculated using the sandwich package from R Studio.

The following variables were used as independent variables:

· presence/absence of sexual experience,

· sexual health literacy rate,

· false knowledge rate,

· risk perception rate,

· perceived severity (`Shame' and `Terrible consequences'),

· perceived susceptibility,

· knowing someone with STDs.

The following control variables were also added to the model:

· age,

· gender,

· mother's education.

Initially, models with the dependent variable were compiled separately with each independent variable. Further, all variables that showed a significant effect (sexual health literacy, risk perception) were combined into one model. For each model, R2, Adjusted R2, Residual Standard Error, Akaike's Information Criteria (AIC), F Statistic was calculated, and Estimates from the sandwich package (R Studio package).

The second model focuses on sexual behaviour (whether the teenager used any contraception to prevent unplanned pregnancy during the last sexual contact) as its dependent variable, which has two levels: 1 for the case when a teenager would have used contraception to prevent unplanned pregnancy during the last sexual contact, and 0 for the situation of not using a contraception. For this model, only the responses of students who already had sex (671 respondents) were used.

To build the model, binomial linear regression was used, and the odds ratio with 95% confidence intervals was calculated. For this, packages from R Studio were used. The following variables were used as independent variables:

· sexual health literacy rate,

· false knowledge rate,

· risk perception rate,

· perceived severity (`Shame' and `Terrible consequences'),

· perceived susceptibility,

· knowing someone with STDs,

· self-efficacy of condom use.

The following variables were included in the model as control ones:

· age,

· gender,

· mother's education.

Initially, models with the dependent variable were compiled separately with each independent variable. Further, all variables that showed a significant effect (self-efficacy, risk perception) were combined into one model. For each model Log Likelihood (LL) and Akaike's Information Criteria (AIC) were utilized.



3. Results

3.1 Descriptive Statistics

The average age of adolescents participating in the study is 18 years old (from 17 years to 21 years), 37% of respondents are girls (Table 1). Most sexually active adolescents (Table 1, 67%) used a contraception during their last sexual contact (Table 1, 81%).

Regarding sexual literacy, the average indicator of knowledge about the effectiveness of different methods of contraception is ~7 out of 13 maximum possible (Table 1). The average false knowledge rate in adolescents is relatively low and amounts to ~1 out of 13 of the maximum possible (Table 1).

81% of adolescents who participated in the study reported that they did not know people with STDs, and only 19% acknowledged that they knew 1 or more people with sexually transmitted diseases (Table 1).

The average score of self-efficacy of condom use is ~4, thus, adolescents are generally more confident about their condom use efficiency (Table 1).

Concerning risk perception, the average indicator shows that adolescents tend to assess the risks of having an unwanted pregnancy with unprotected sex as “50 to 50” (Table 1).

As for the consequences for a teenager in case of unplanned pregnancy, on average, adolescents do not consider unplanned pregnancy a disgrace for themselves or their family (Table 1). In situations where an unwanted pregnancy is described as a terrible event and adolescents planning an abortion, adolescents are less confident and, on average, choose the answer (3) “I don't agree nor disagree (Table 1).



Table 2 demonstrates the difference between male and female survey respondents regarding their mother's education: mothers of 51% of male and 33% of female participants possess a higher education degree (Table 2). What is more, the level of sexual literacy is higher among female (7.58) than among male teenagers (6.52) (Table 2, out of 13 maximum possible). Regarding the use of contraception for preventing unwanted pregnancy, most male adolescents themselves took responsibility for use of contraceptives - 38% of the research participants (Table 2). Among females, this indicator equals to only 6%, and 30% of girls admitted that it was their partner who took this responsibility (Table 2). Among male and female teenagers, 18% and 22%, respectively, did not use any contraception, which would have protected them from unwanted pregnancy during the last sexual contact (Table 2). Statistical tests (t-test and chi-squared test) were also conducted. Gender turned out to be significant for all variables listed in Table 2 (mother's education, sexual health literacy, sexual behaviour).

Table 2

Chart 1 shows that most adolescents indicated that the likelihood that they will get STDs is very small or practically non-existent. Moreover, 536 of them also indicated that they did not know a single person with STDs. Nevertheless, the difference was insignificant (p-value = 0.14).



3.2 Regression models results

Self-efficacy of condom use

A high-risk assessment regarding pregnancy as a consequence of an unprotected sexual intercourse demonstrates a strong association with using condoms by adolescents: the higher the risk assessment is, the higher the probability that a teenager's self-efficacy of using a condom is high (Table 4, model 1, Est. = 0.082***).

What is more, the higher the level of sexual literacy a teenager has (he/she holds correct information about a particular type of contraception and its effectiveness), the greater the probability that his/her self-efficacy of condom use is high (Table 3, Model 1, Est. = 0.036***).

Table 3

Control variables (age, gender, mother's education level) did not show significant associations (App. 2, model 1). The results of control variables' effect on condom use reflect that male and female teenagers of different ages equally evaluate self-efficacy. Besides, adolescents whose mothers had higher education and those whose mothers with lower education, or an incomplete higher education, also do not differ in assessing their self-efficacy.

Interactive effects between gender and sexual literacy, gender and risk perception were tested. As it can be seen from Table 3 (model 2), the perception of risks has a different effect on the self-efficacy among males and females. Chart 2 shows the difference between males (1) and females (0), conveying that the higher the perception of risks is, the higher the self-efficacy of condom use. Nevertheless, from Chart 2 it can be concluded that females tend to rate the risks higher than male adolescents do, and, in fact, evaluate their self-efficacy higher.

The interactive effects of sexual literacy and gender were also tested. Table 3 (model 3) reveals that the differences between the sexual literacy among male and female teenagers are not statistically significant for the self-efficacy of condom use.

Comparing all three models with each other, the highest relative quality is provided by model 2, which includes not only perceived risks and sexual health literacy, but also an interactive effect of gender and risk perception (Table 3, AIC=1975.450).

App. 2 presents the results of logistic regression for the dependent variable of self-efficacy and all the independent variables. The first thing to note is the fact that the presence of false knowledge about different types of contraception and their effectiveness did not show a significant association with self-efficacy (App. 2, model 3).

Furthermore, perceived severity (App. 2, model 4, model 5), perceived susceptibility (App. 2, model 6), knowing someone with STDs (App. 2, model 7), and sexual experience (App. 2, model 2) did not show a significant association with the dependent variable. Thus, perceived severity, knowing someone with STD, perceived susceptibility, and the presence/absence of sexual experience are not significantly related to how teenagers evaluate their self-efficacy of condom use.

Sexual Behaviour

Table 4 illustrates that, taken constant values ??of other factors in consideration for male participants of the study, the odds ratio of using contraceptives is 2.2, thus, males use contraceptives more often than females.

The odds ratio of using contraception will increases by 2.02 times with an increase in risk assessment by one with constant values ??of other variables in the model (Table 4). If an increase in the risk perception by one triggers the odds ratio of using contraception to enhance increases 2.02 times, then with a value of risk = 5, the odds ratio of using contraception would be 33.63 with constant values of the other variables (Table 4). This means that if a male admits the risk of an unplanned pregnancy in the case of a sexual intercourse without a condom, then he has a better plausibility of using contraception during sexual contact.

The odds ratio of using contraception will increases by 1.81 with an increase in self-efficacy by one with constant values ??of other variables in the model (Table 4). That is, the higher a teenager's self-efficacy of condom use is, the higher the chance that the teenager uses contraception gets. If an increase in self-efficacy by one increases the odds ratio of using contraception by 1.81 times, then an increase in self-efficacy by 5 increases the odds ratio of using contraception by 19.43 with constant values ??of the other variables (Table 4).



Table 4

Control variables of a participant's age and his/her maternal education did not show a significant effect on the use of contraception in the last sexual encounter (App. 4, see in the appendix). Thus, adolescents of different ages and teenagers whose mothers had higher education and those whose mothers had lower education or incomplete higher education do not differ in the use of contraception during their last sex. Interactive effects were tested between gender and the self-efficacy of condom use, gender and risk perception. Nevertheless, for both cases, there were no significant differences between females and males (App. 3).

App. 4 presents the results of regression for the self-efficacy of condom use during the last sexual contact and all the independent variables. The sexual health literacy (App. 4, model 4) and false knowledge (App. 4, model 5) did not produce significant results. Thus, sexual literacy and the presence of false knowledge are not associated with the use of contraception during the last sexual encounter.

As for perceived severity (App. 4, model 2, model 3), perceived susceptibility (App. 4, model 6), and knowing someone with STD (App. 4, model 7), they did not show a significant association with the dependent variable. Accordingly, perceived severity, being familiar with someone with STD, and perceived susceptibility are not significantly related to the use of contraception during the last intercourse. Control variables (age, gender, maternal education level) also did not show significant associations (App. 4, model 1). Females and males would be equally likely to use contraception during their last sex, and this fact is neither related to the age of the teenager nor their mother's obtained degree of education.



4.Discussion

The relationship between sexual health literacy and self-efficacy of condom use and safe sexual behaviour

The purpose of this study was to describe how sexual literacy is associated with self-efficacy of condom use and safe sexual behaviour. In the explored literature, the relationship between sexual literacy and safe sexual behaviour is rather mixed (Kotchick et al. 2001): some studies discover a positive relationship between the two factors (Mahat and Scoloveno 2010; Kayiki and Forste 2011; Swenson et al. 2010), while others find no connection (Levinson 1995; Lou and Chen 2009; Rock 2005). For sexually active adolescents who participated in this study, sexual health literacy did not show a significant association with condom use during their last sexual intercourse. However, for a sample that included not only sexually active adolescents but inexperienced teenagers too, sexual literacy was positively associated with self-efficacy of using condoms.

The Health Belief Model suggests that knowledge does not directly affect human behaviour (Rosenstock, Strecher, and Becker 1988, 1988). This model assumes that it is not enough for a person to simply have the necessary knowledge to complete a particular preventive action, but a person should also feel that he/she is susceptible to this disease, plus, he/she must have a high self-efficacy in taking preventive measures, and, respectively, he/she must understand that he/she will be able to take preventive action (Rosenstock, Strecher, and Becker 1988, 1988). Thus, besides sexual literacy and risky sexual behaviour, there are several other factors that also perform a significant role and influence human behaviour. A number of researchers support the point of view that holding certain knowledge is not enough: social intentions and perceptions (Langille et al. 1998), moral judgment (Hubbs-Tait and Garmon 1995), perceived knowledge (Rock 2005) are also important.

Misconceptions and false knowledge about contraception, STDs, and HIV/AID

Previous studies have shown that false knowledge is present in older adolescents who already had sexual experience (Crosby and Yarber 2001); in this study, the presence of false knowledge did not show significant results for both models. In general, adolescents from colleges in St. Petersburg have a relatively low level of false knowledge regarding various types of contraception and their effectiveness, as an average score is 1/13 of incorrect answers about contraception's effectiveness. Nevertheless, the presence of false knowledge can influence risk assessment: false knowledge about effective contraceptive methods or STDs may reduce perceptions of risk and one's vulnerability, which can lead to a teenager becoming involved in risky sexual behaviour and seeking medical attention later (Bhattacharya, Cleland, and Holland 2000). Thus, like sexual literacy, the presence of false knowledge is not directly related to safe behaviour and may have a major impact not on the sexual behaviour itself, but on other factors like the perception of risks, perception of one's vulnerability, and so on.

Risk perception and self-efficacy as the predictors of safer sexual behaviour

In both cases, the relationship between risk assessment and perceived self-efficacy of condom use and safe sexual behaviour is clearly seen. This conclusion is supported by other studies (Wiener, Battles, and Wood 2007; Slonim-Nevo and Mukuka 2005), which have proved that the higher adolescents rate the risks of unwanted pregnancy is, the greater the probability that they will use contraception becomes. Moreover, the higher a teenager estimates the risks, the more he/she is confident that he/she will use a condom. Wiener, Battles, and Wood (2007) suggest that a teenager's confidence that he/she will be able to use a condom during sex is a key factor that affects safe sexual behaviour. In the performed study, adolescents who rated the self-efficacy of condom use higher were more likely to use contraception during their last sexual contact. This relationship between self-efficacy and safe sexual behaviour is explained by Health Belief Model: in order to take preventive actions, a person must be confident in the self-efficacy of implementing preventative measures (Rosenstock, Strecher, and Becker 1988, 1994), in this case, to use contraception.

Perceived susceptibility and knowing someone with STD as predictors of safer sexual behaviour

Although many studies have shown that being familiar with a person with an STD or communicating with someone who would have died because of HIV is associated with more frequent use of condoms (Ekanem et al. 2005; Kayiki and Forste 2011), in this study, those familiar with STDs were not associated with neither safe sexual behaviour nor self-efficacy of condom use, moreover perceived susceptibility also did not show significant results. Almost 81% of surveyed students did not know a single person who admit possessing sexually transmitted diseases, and most of them knew 1-2 people. In reality there may be much more people with STDs, but because of the stigma regarding STDs, people prefer not to tell people and even friends about their status (Cunningham et al. 2009). 536 adolescents who mentioned that they did not know people with STDs, also mentioned that they are not susceptible to sexually transmitted diseases. A feeling of false security may develop around a teenager an erroneous impression that his/her friends and other peers do not suffer from STDs. The influence of adolescents was well demonstrated in a study by Brown, DiClemente, and Reynolds (1991), who showed that adolescents who were confident that their peers used condoms were more likely to use condoms themselves in comparison with those who did not believe in the STDs' spread.

Perceived severity as a predictor of a safer sexual behaviour

How seriously people evaluate their health outcomes is an important predictor of probability that a person would take preventative measures (Rosenstock, Strecher, and Becker 1988, 1994). In this research, perceived severity was not studied in terms of serious health consequences, but as an impact of negative social consequences that could occur in the event of an unplanned pregnancy. Neither for a model that predicts the perceived self-efficacy of condom use nor for a model that predicts safe sex behaviour, perceived severity regarding the social consequences for a teenager did not turn out to be significant.

These results reveal that some components of HBM were indeed insignificant for the use of contraception, however, there may be other factors that influence teenagers' behaviour, and they may be even stronger than the elements of HBM. Researchers explain that cognitive models are, in principle, limited in their ability to explain safe sexual behaviour, especially if it is related to HIV/AIDS (Fisher and Fisher 2000). Researchers also note the important role of sociocultural factors, such as condom promotion, for safe sexual behaviour (Wyatt et al., 1999). HBM could predict more differences in sexual behaviour if a factor such as a peer influence would be added to it (Brown, DiClemente, and Reynolds 1991). Thus, it is possible that after inclusion of sociocultural factors in the model, as well as the perception of norms by peers, the model may show more differences and will be more effective in predicting safe sexual behaviour.

Social-demographic factors

Variables that relate to the family structure, such as the mother's education, which was used in this study, are often statistically controlled because of a present correlation with other variables (Kotchick et al. 2001). Thus, it makes it difficult to draw any conclusions. In this study, the mother's education did not show significant results for the self-efficacy of condom use and safer sexual behaviour. According to the researchers, adolescents aged 15-19 years are much more likely to know about AIDS compared to adolescents 10-14 years old (Khan 2002). In this study, the age of the research participants (aged 17-21) did not show differences in self-efficacy of condom use and safer sex behaviour.