A Network Approach to Sexual Self-Concept: Analyzing the Multidimensional Sexual Self-Concept Questionnaire

Master Thesis

A Network Approach to Sexual Self-Concept: Analyzing the Multidimensional Sexual Self-Concept Questionnaire

G. Servaite s2454416 MSc Clinical Psychology Supervisor: E.I. Fried, Ph.D. Institute of Psychology University Leiden 15-07-2020

Table of Contents

1. Introduction ... 4

1.1 Sexual Self-Concept ... 5

1.2 The development of sexual self-concept ... 7

1.3 Factors that influence sexual self-concept ... 8

1.4 Gender ... 9

1.5 Sexual self-concept measurement ... 11

1.6 The MSSCQ ... 12

1.7 Research questions and hypotheses ... 19

2. Methods ... 20

2.1 Participants ... 20

2.2 Measures/Instruments ... 21

2.3 Research design and statistical methods ... 22

3. Results ... 24

3.1 Demographic characteristics ... 24

3.2 Network analysis of the sexual self-concept ... 24

1.1 Network structure of sexual self-concept across gender ... 26

1.2 Network structure of sexual self-concept across age groups ... 28

4. Discussion ... 30

4.1 Sexual self-concept connectedness and centrality ... 31

4.2 Sexual self-concept across genders ... 34

4.3 Sexual self-concept across age groups ... 35

4.4 Limitations ... 36

4.5 Conclusions ... 37

Appendix 1 ... 38

Appendix 2 ... 39

Abstract

Sexual self-concept has been found to be crucial for the positive approach to human sexuality. However, the lack of consensus within the sexual self-concept research field suggests that current portrayal of sexual self-concept may be incomplete and should be further explored. This study investigates the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ) in a sample of 12,449 respondents using network analysis. The research questions sought to examine how the aspects of sexual self-concept are situated in a network and whether the structure of the networks differs between genders and different age groups. Regularized partial correlations were estimated among 20 aspects of MSSCQ. The found network of sexual self-concept was strongly interconnected with both positive and negative partial correlations. The strongest connections in the sexual self-concept network were between sexual satisfaction and sexual depression, sexual anxiety and sexual depression, sexual problem management and sexual problem self-blame, sexual preoccupation and sexual motivation, and sexual anxiety and sexual fear. Sexual depression, sexual anxiety and sexual motivation were the most interconnected aspects in the network, presumably playing an important role in one’s sexual concept. The structure of the sexual self-concept network was surprisingly similar across genders and age groups. This is the first study to examine the network structure of sexual self-concept, and one of very few studies utilizing the network approach on such a large sample size. The applied psychological network methodology uncovers the complexity of sexual self-concept and allows for a deeper understanding of the interactions that take place.

1. Introduction

Sexual health is essential to one’s quality of life and general well-being. Many different factors - psychological, biological, social, political, cultural, religious, spiritual – shape and affect it, thus illustrating its complexity (Do, Khuat, & Nguyen, 2018). Discussing sexual health raises the question what it means to be sexual and, therefore, it requires an understanding of human sexual behavior and a positive approach to human sexuality to ensure it (WHO, 2006). Paradoxically, although sexuality is an integral part of one’s identity, it is especially difficult for individuals to explore and express this part of self (McKenna, Green, & Smith, 2001).

Defining sexual health has been closely connected and shaped by political, social and other historical events – the sexual revolution, ongoing debates about reproductive rights, the gay rights movement and many others (Edwards & Coleman, 2004). Initially not recognized as a distinct scientific concept, sexual health was regarded as a subset of reproductive health. It was also debated whether objectively defining sexual health would mean establishing norms for it, labeling some behavior as “healthy”, and other, not fitting the norms as “unhealthy”, therefore becoming a tool to exclude or label those that did not fit the objective definition (Edwards & Coleman, 2004). On the other hand, defining sexual health was a crucial step in order to objectively measure it and, ultimately, protect it.

Most research on sexual health over the years seemed to hold a narrow perspective and focus on negative and problematic outcomes of sexuality, especially in terms of adolescent sexuality. Numerous studies were conducted on sexual risk taking, condom use, sexually transmitted infections and unplanned pregnancy (Rostosky, Dekhtyar, Cupp, & Anderman, 2008). Over the last decades, however, research on sexuality has been notably expanding and new, previously neglected areas have received increased attention. With the revolution of positive psychology came a focus on how healthy and normal individuals can get the most from life (Marsh & Craven, 2006). This naturally involved sexuality as well – effort was made to understand sexual health in a more positive light. This observed paradigm shift in the sexual health field introduced a more holistic approach, which aimed at integrating somatic, emotional, intellectual and social aspects of sexual well-being (Deutsch, Hoffman, & Wilcox, 2014; Edwards & Coleman, 2004). According to the World Health Organization’s (2006) latest revised definition of sexual health, it extends beyond “the absence of disease, dysfunction or infirmity”, and highlights the interplay of physical, emotional, social and mental well-being. Also, it recognizes pleasure as a viable

component as well as respect, protection and fulfillment of people’s sexual rights. This recent shift from a negative illness-based towards a more positive, health-promoting framework has created the possibility to examine what constitutes sexuality, which factors influence it and how it is experienced (Blunt, 2012). Expanding the research scope has allowed to go beyond sexual problem solving into topics such as sexual satisfaction (e.g. Fallis, Rehman, & Purdon, 2014), sexual pleasure (e.g. Barnett & Melugin, 2016), sexual self-efficacy (e.g. Sullivan, McPartland, Price, Cruza-Guet, & Swan, 2013) and sexual self-concept (e.g. Anticevic, Jokic‐Begic, & Britvic, 2017), which are gaining momentum in current research.

1.1. Sexual Self-Concept

Self-concept is considered to be one of the oldest areas of research in the social sciences (Marsh & Craven, 2006). It dates back to William James’s 1908 introductory textbook of psychology and his study of the self, which is considered to be the source of the concept. Self-concept is defined as the individual’s belief about themselves, including their self attributes (Baumeister, 1999). However, the predictive value of global self-concept is being increasingly questioned. A number of researchers have pointed out that it could be improved by narrowing the measurement to a certain, more specific aspect of self-concept (Swann, Chang-Schneider, & McClarty, 2007; Marsh & Craven, 2006). That is precisely what the multidimensional perspective postulates: Instead of relying on the global component of self-concept to predict an outcome, one should undertake a domain-specific assessment of self-concept to predict an outcome in that domain (Marsh & Craven, 2006). This very much applies to sexual self-concept – researchers find the measures that define it more narrowly to typically demonstrate better reliability and validity than the measures that define the concept more broadly (Zeanah & Schwarz, 1996).

Sexual self-concept refers to a person’s view of themselves as sexual beings. It is a multidimensional construct that consists of interpersonal and intrapersonal dimensions and includes both positive and negative evaluations of oneself as well as sexual experiences with others (O’Sullivan, Meyer-Bahlburg, & McKeague, 2006). It involves evaluating one’s personal sexual feelings, thoughts and actions (Winter, 1988) and identifying one’s sexual desires and tendencies (Ziaei, Khoei, Salehi, & Farajzadegan, 2013). Moreover, it helps in organizing and making sense of sexual experiences, raising awareness, self-evaluation and acquiring understanding of one’s sexual life (O’Sullivan et al., 2006). Sexual self-concept is an active, dynamic construct that forms

over time through individual interpretations of sexual experiences and external feedback from others (Deutsch et al., 2014). Different aspects of one’s sexuality interact with the sexual environment one is in, forming a reciprocal relationship with sexual self-concept (Deutsch et al., 2014). However, this reciprocal approach is still overshadowed by the more straightforward, unidirectional approach which views sexual self-concept as the antecedent to sexual behavior (Hensel, Fortenberry, O’Sullivan, & Orr, 2011). There are other terms used synonymously to sexual concept in academic literature, such as sexual selfhood, sexual subjectivity, sexual self-esteem or sexual self-schema (Deutsch et al., 2014; Zeanah & Schwarz, 1996). Despite this, sexual self-concept remains the most commonly used with the largest body of research (Deutsch et al., 2014).

Sexual self-concept has been found to be important for the positive approach to human sexuality. Individuals with higher sexual self-concept scores are reportedly found to be more sexually satisfied in their relationships, which suggests a direct relationship between sexual self-concept and satisfaction (Blunt, 2012). It also serves as a protective factor in adolescent sexual health (Pai, Lee, & Yen, 2012). Adolescents’ positive sexual self-concept mediates sexual risk knowledge and sexual self-efficacy, and thereby enhances their ability to employ knowledge of sexual risk into action and preserve their sexual well-being (Rostosky et al., 2008). Sexual self-concept also strongly correlates with adolescent girls’ sexual health behavioral intentions (Pai et al., 2012) as well as adolescents’ contraceptive behavior (Winter, 1988).

A lack of consensus has been observed within the sexual self-concept research field. Sexual self-concept studies have been predominantly exploratory. Researchers have developed sexual self-concept models by performing factor analysis on items that they themselves conceptualize to relate to sexual self-concept (Ho et al., 2017). This leads to a collection of partly overlapping, nonetheless different sexual self-concept models. Although the conceptual definition of sexual self-concept is commonly agreed upon and many sexual self-concept models overlap in some aspects, the lack of consensus between the models suggests that the true factor structure of sexual self-concept is yet to be understood and the current portrayal of sexual self-concept may be incomplete (Deutsch et al., 2014). This might be due to sexual self-concept not being linked to any specific theoretical framework (Deutsch et al., 2014). Therefore, considering the importance of sexual self-concept in the field of sexual health, it is necessary to collect more information about the above-mentioned concept, its specificity, core components and the interactions between them.

1.2. The development of sexual self-concept

Sexual self-concept develops together with sexual experiences, and becomes, over time, less context dependent and more integrated into the global self-concept (Deutsch et al., 2014). Although the beginning of active sexuality formation is considered to take place during puberty, the continuity of sexual development reaches into developmental stages as early as infancy. Preverbal experiences enable the development of verbal capacities and, together with early experiences of pleasure, comfort, as well as guilt, anxiety and ambivalence, they form the basis of sexual complexity (Gagnon & Simon, 1974). The post-verbal years until puberty are heavily influenced by parents’ reactions to the child’s behavior which can be perceived as sexual as well as encouragement of conventional gender identity (Gagnon & Simon, 1974).

One of the primary developmental tasks during adolescence is the development of a positive sexual concept (Rostosky et al., 2008). Each dimension of which the sexual self-concept consists evolves in its own unique way throughout adolescence, and this active and dynamic organizational process eventually leads to an internalized coherent construct (Hensel et al., 2011; Deutsch et al., 2014). Early adolescents start to experience physical and emotional changes that accompany puberty. They become aware of their sexual attraction and begin to view themselves as capable of evoking sexual interest in others (O’Sullivan et al., 2006). Identifying and integrating these experiences and moving towards a growing sense of self is crucial at this point of adolescence (Erikson, 1950). In the early stages of adolescence, sexual self-concept is less differentiated and more dependent on different contexts due to a lack of personal sexual experiences and yet not fully developed cognitive functions (Deutsch et al., 2014). Although early adolescents are considered to be sexually not yet active, with first intercourses more likely to occur during middle to late adolescent years (aged between 15 and 21), there is a considerable range of pre-coital sexual experiences that precede the first intercourse and which are often overlooked (O’Sullivan et al., 2006). They include the experiences, behaviors and emotions that the sexual self becomes linked to, thereby expanding one’s “sexual repertoire” and its associated meanings and leading to a recalibrated sexual self-concept (Hensel et al., 2011). What is more, peer networks become relevant in early adolescence and provide the social contexts for early romantic and sexual encounters (O’Sullivan et al., 2006). Social norms and peer feedback - scrutiny or approval - guide sexual behavior. It is therefore important to take relational factors into account when assessing

adolescents’ sexual self-concept and sexual behavior (Rostosky et al., 2008; Tolman & McClelland, 2011).

Sexuality issues become of particular importance in the late adolescent age group, also defined as college-age, around 18 to 22 years (Zeanah & Schwarz, 1996). By that time, most individuals have personal and social contexts to draw from and make self-evaluations in the sexual domain. Some theorists distinguish emerging adulthood, a period from the late teens through the twenties, distinct from adolescence and young adulthood and consider it important to sexuality development (Arnett, 2000). It refers to the time when the dependency characterizing the childhood and teenage years is left behind and the responsibilities of adulthood are yet to come. This creates the possibility to explore different life directions and roles, thus becoming the time for sexual experimentation, cohabitating and exploring deeper levels of intimacy, all of which further refine one’s sense of sexual self (Arnett, 2000). Interestingly, however, this concept only exists in developed societies, where young people are allowed prolonged exploration - both socially and economically (Arnett, 2000).

Overall, the previously mentioned shift towards an expanded understanding of what constitutes sexual health does not bypass the field of adolescent sexuality. Researchers are increasingly advocating for adolescent sexuality to be unlinked from pathology and dangerous outcomes, and for adolescents to be viewed as actively making decisions regarding their positively motivated sexuality (Tolman & McClelland, 2011). This emerging normative approach to adolescent sexuality allows for viewing it as an expected and natural part of one’s development. A better understanding of adolescents’ sexual self-concept can therefore contribute to enabling young people in their journey of becoming sexually healthy adults.

1.3. Factors that influence sexual self-concept

The literature on sexual self-concept can be divided into two research areas: those focused on designing instruments to assess sexual self-concept and those examining relationships between sexual self-concept and other variables (Deutsch et al., 2014). The collaboration between the two areas could potentially bring mutually beneficial results, both advancing the measurement instruments as well as broadening our understanding of sexual self-concept. A number of factors have been shown to affect sexual self-concept and can be summed up in three categories: biological, social and psychological (Potki et al., 2017). Biological factors refer to age, gender,

marital status, race and disability, social factors include the role of parents, peers and the media, and the psychological category involves body image, sexual abuse in childhood and mental health history (Potki et al., 2017).

Several aspects deserve a closer look. Firstly, sexual self-concept has been found to increase (i.e. become more positive) with age (Winter, 1988). Over time and with acquired sexual experience, sexual anxiety is reduced and comfort built, thus creating a positive environment for subsequent sexual behavior (O’Sullivan et al., 2006). Obtained sexual experience also boosts one’s sexual self-confidence and, consequently, reduces negativity towards sexual issues (Potki et al., 2017). However, the relationship between age and sexual self-concept is more complex and not a strictly linear one. While broadening sexual horizons and acquiring new partners, concerns about reputation and competence might reemerge, which can lead to increased sexual anxiety (Hensel et al., 2011). Secondly, sexual self-concept has been connected to marital status. Unmarried individuals report higher levels of sexual anxiety, sexual self-monitoring and sexual fear, and lower levels of sexual assertiveness, compared to married individuals (Snell, Jr Snell, & editor, 2001). Divorced individuals report higher levels of sexual self-efficacy, greater motivation to avoid high-risk sexual behavior and higher levels of sexual fear, while re-married individuals report lower levels of sexual anxiety, sexual self-monitoring, sexual fear and motivation to avoid risky sex, and higher levels of sexual assertiveness (Snell et al., 2001). Thirdly, the media has a particularly strong effect on adolescents and adults’ sexual self-concepts. Although having secured its role in adolescents’ sexual education, it has also been found to damage women’s sexual self-concept (Aubrey, 2007). Frequent exposure to media sources of sexually related messages has been found to be linked to greater acceptance of sexual stereotypes and casual attitudes about sex as well as distorted expectations towards sexuality and sexual experiences (Ward, 2003; Tolman & McClelland, 2011). For example, intense consumption of soap operas has been linked to over-estimating the frequency of people’s sexual activities and the difficulties in maintaining romantic relationships (Aubrey, 2007).

1.4. Gender

An important factor to touch upon in the context of sexual self-concept is gender. It was found to be a fundamental as well as controversial aspect of how people perceive their sexuality and its development.Some researchers highlight significant gender differences in how sexual self-concept

operates (Deutsch et al., 2014; Breakwell & Millward, 1997). Men’s overall sexual self-concept scores have been found to be higher than those of women (Deutsch et al., 2014). Men also perceive themselves as more sexually responsive, deviant and experienced whereas women perceive themselves as more sexually attractive and romantic (Garcia & Carrigan, 1998; Garcia, 1999). Research findings suggest that women’s sexual self-concept is not as stable as men’s, with situational, interpersonal and contextual factors more likely to influence it. Women’s sexual self-concept was found to undergo major changes during (and after) life-changing events such as giving birth (Ho et al., 2017).

Moreover, researchers highlight the impact of internalized societal sexual roles on individuals’ sexual self-concept. Men and women learn to perceive themselves differently as sexual beings through different feedback they receive: more positive feedback when adhering to hegemonic norms of femininity and masculinity, or criticism when violating them (Deutsch et al., 2014). Therefore, the social construction of gender leads to the social construction of appropriate sexual behavior, which differs between men and women (Butler, 2004; Lorber, 2006). This can be especially problematic for adolescents’ sexual development as gender plays an important role in the formation of their sexual identity. Femininity norms can negatively impact girls’ psychosocial development, consequently leading to personality changes, body image dissatisfaction and other detrimental issues (Tolman, Striepe, & Harmon, 2003).

Furthermore, although most research in the field of sexual self-concept has focused on women, their sexuality has nevertheless been mostly studied from a male perspective, based on what was known about male sexuality (Ho et al., 2017). Women were typically viewed as sexually passive and were unable to freely express their sexuality, with societal limitations dictating what is appropriate. In the recent decades, however, researchers have started to approach women’s sexuality from a more holistic perspective by incorporating women’s unique experiences to better capture and depict them (Mollen & Stabb, 2010). For example, Nicolson and Burr (2003) found that women tend to have a stronger desire to experience orgasm for their male partner’s sake rather than for themselves. It has also been found that women sometimes struggle to recognize the physiological cues which indicate their arousal, which implies that their subjective interpretation may be inconsistent with their physiological arousal (Basson, 2002; Mollen & Stabb, 2010).

On the other hand, recent studies suggest a narrowing gap between gender differences, with the reason being that, at least in part, women’s sexuality is increasingly closing in on men’s

(McCabe, Tanner, & Heiman, 2010). Some theorists claim that there are more differences amongst women and amongst men than between them (Rutter & Schwartz, 2012). Both genders relate to the same basic dimensions when rating their sexual selves (Hill, 2007). The more personal the area of assessed sexuality is, the more similarities are found between men and women (Deutsch et al., 2014). People tend to draw on stereotypical gender norms when talking about sexuality in abstract, but, when considering their own sexual experiences and relationships, they are more likely to violate the gender-specific patterns (McCabe, et al., 2010). The effect of the sociocultural environment and gender expectations could thus be rather limited, and the sexual differences exaggerated by both the environment and society.

1.5. Sexual self-concept measurement

There have been numerous attempts to conceptualize and measure sexual self-concept. Scales differ not only in terms of the items and dimensions they encompass but in terms of the target groups as well. Instruments for older adolescents and adults involve different item content, format and wording due to a likely wider range of sexual experiences to refer to. Therefore, different sexual self-concept models might be applicable to different age groups. The scales are often gender-specific due to supposed varying trajectories of sexual development (O’Sullivan et al, 2006). Vickberg and Deaux (2005) propose The Women’s Sexual Self-Concept Scale (WSSCS), a valid and reliable measure for women’s sexual self-concept assessment, while O’Sullivan et al. (2006) offer a three-dimensional model of sexual self-concept, consisting of sexual agency, sexual arousability and negative sexual affect, developed specifically for ethnically diverse early adolescent girls. Buzwell and Rosenthal’s (1996) empirically supported sexual selfhood model for both genders includes sexual self-esteem, sexual self-efficacy, arousal, exploration, anxiety and commitment. Considerable critique was raised over the widely used Andersen and Cyranowski (1994; 1999) sexual self-schema scales, where women’s scales reportedly include more negatively valued items than men’s scales, and the traits in the scales are strongly related to gender stereotypes (Hill, 2007).

Snell and Papini (1989) created a measure of esteem, depression and sexual-preoccupation that later was expanded into the Sexual Awareness Questionnaire (SAQ), measuring four personality tendencies of sexual self-concept: sexual monitoring, sexual preoccupation, sexual consciousness and sexual assertiveness (Snell, Fisher, & Miller, 1991). This led to the

Multidimensional Sexuality Questionnaire (MSQ), which measures 12 aspects of the sexual self-concept, including several new subscales designed to assess control-related aspects (Snell, Fisher, & Walters, 1993). This was eventually expanded into a more global and comprehensive measure of sexual self-concept that is used to this day and analyzed in the current study: the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ; Snell, 1998).

1.6. The MSSCQ

The MSSCQ is considered to be one of the most practical and widely utilized instruments and is composed of 20 aspects that cover cognitive features (e.g. sexual self-schemata), affective components (e.g. sexual depression) and motivational factors (e.g. sexual motivation) (Ziaei et al., 2013). The scale is a self-report instrument, designed to measure 20 psychological aspects of human sexuality, each of which is discussed below. The measurement of sexual self-concept is not based on one total score but rather on individual scores of each sexuality-related construct (Ziaei et al., 2013; Fisher, Davis, & Yarber, 2011). It is thus important to examine in detail each individual aspect that constitutes sexual self-concept.

1.6.1. Sexual anxiety

Sexual anxiety is defined as the tendency to feel worry, tension, uneasiness and discomfort about one’s sexual life aspects (Snell et al., 2001). It is one of the most commonly reported problems and contributing factors to a variety of sexual dysfunctions (Rosen, Taylor, Leiblum, & Bachmann, 1993; Ingersoll & Dyson, 2011). A decline in sexual anxiety has been found to be linked to a growth in sexual self-esteem (Hensel et al., 2011). Studies also show a positive correlation between sexual anxiety and sexual depression (Offman & Matheson, 2005). Moreover, sexual anxiety has also been shown to decline with age and gained sexual experience. Women report more sexual anxiety compared to men (Vanwesenbeeck, Bekker, & van Lenning, 1998; Snell et al., 2001).

1.6.2. Sexual self-efficacy

Sexual self-efficacy is defined as the belief that one has the ability to deal effectively with the sexual aspects of oneself (Snell et al., 2001). It is understood as an individual’s level of confidence in sexual situations including their ability to suggest and use contraception with their partner (Norton, Smith, Magriples, & Kershaw, 2016). Some theorists further distinguish between

situational sexual self-efficacy and resistive sexual self-efficacy. Situational sexual self-efficacy reflects one’s confidence in their ability to manage potentially risky sexual situations, while resistive sexual self-efficacy focuses on lack of engagement and the perceived ability to take initiative, be responsible and resistive to unwanted sexual activity (Rostosky et al., 2008; Deutsch et al., 2014). Higher sexual self-efficacy has been linked to condom use, lower sexual risk taking and higher sexual self-esteem (Rosenthal, Moore, & Flynn, 1991; Rostosky et al., 2008). Sexual self-efficacy is reportedly higher among women (Rostosky et al., 2008). However, past research has focused on sexual self-efficacy among women while neglecting men and the shared responsibility of sexual decision-making (Rostosky et al., 2008).

1.6.3. Sexual consciousness

Sexual consciousness is defined as the tendency to be aware and reflective about the internal aspects of one’s sexuality (Snell et al., 1991; Snell et al., 2001). Individuals with high sexual consciousness pay more attention to their sexuality, inspect and examine their sexual desires, motivations and thoughts, tend to practice safe sex and approach sex in an interpersonal manner (Snell et al., 1991). Sexual consciousness was found to be positively correlated with an internal locus-of-control and negatively correlated with sexual guilt, sexual anxiety and powerful-other sexual control among women (Snell et al., 1991). Sexual consciousness was also positively correlated with sexual self-esteem, sexual preoccupation and sexual satisfaction, while being negatively correlated with sexual depression (Snell et al., 1991).

1.6.4. Motivation to avoid risky sex

Motivation to avoid risky sex is defined as the desire to avoid unhealthy patterns of risky sexual behaviors (e.g. unprotected sexual behavior) (Snell et al., 2001). It relates to an individual’s willingness to avoid sexual behavior that may compromise health and well-being. Women have been found to be more motivated to avoid risky sex in contrast to men (Snell et al., 2001). However, interestingly, motivation to avoid risky sex does not necessarily determine using contraception. Contraceptive use is a planned action that requires anticipating the probability of sexual activity, and different factors, such as sexual anxiety or fear, might interfere with this anticipation and obtaining birth control (Winter, 1988).

1.6.5. Chance/luck sexual control

Chance or luck sexual control is defined as the belief that the sexual aspects of one’s life are determined by chance and luck instead of their own active participation (Snell et al., 2001). It thus refers to an external locus of control. Chance/luck sexual control has been linked to higher sexual risk behavior (Victor & Haruna, 2012). Furthermore, a negative correlation has been found between chance or luck sexual control and sexual assertiveness among women (Snell et al., 1991). Men report higher levels of chance/luck sexual control (Snell et al., 2001).

1.6.6. Sexual preoccupation

Sexual preoccupation is defined as a persistent tendency to think about sex to an excessive degree and to be absorbed with one’s sexual matters (Snell et al., 2001; Snell et al., 1991). These sexual thoughts, fantasies and urges can cause distress and anxiety, and lead to repetitive sexual behaviors such as pornography consumption, excessive masturbation and overly active sexual fantasies (Lee & Forbey, 2010; Noll, Trickett, & Putnam, 2003). Sexual preoccupation has been found to be positively correlated with sexual consciousness, sexual monitoring and sexual assertiveness (Snell et al., 1991; Snell, Fisher, & Schuh, 1992). Interestingly, sexual preoccupation has also been found to be positively correlated with sexual depression among women and sexual self-esteem among men (Snell et al., 1992). Studies have found men to have a higher level of sexual preoccupation (Thurman & Silver, 1997; Snell & Papini, 1989; Snell et al., 1992).

1.6.7. Sexual assertiveness

Sexual assertiveness is defined as the tendency to be highly decisive and self-reliant about the sexual aspects of one’s life (Snell et al., 2001). It involves not only disclosing one’s sexual preferences, but also actively requesting and initiating sexual behaviors – acts that entail much greater interpersonal risks (Menard & Offman, 2009). There is a notable difference between sexual assertiveness in the context of safer sex behavior (asking to use a condom) and sexual assertiveness for the purpose of pleasure and satisfaction –an area which has so far been scarcely researched (Menard & Offman, 2009). Greater sexual assertiveness is directly linked to more reliable contraceptive and STD (sexually transmitted diseases) preventative behaviors (Snell & Wooldridge, 1998; Harlow et al., 1993), higher levels of sexual arousal and desire (Santos-Iglesias, Sierra, & Vallejo-Medina, 2013), less traditional and more positive sexual attitudes (Morokoff et

al., 1997; Santos-Iglesias et al., 2013). Moreover, studies have found sexual assertiveness to be positively correlated with sexual satisfaction and sexual self-esteem and negatively correlated with sexual depression and sexual anxiety (Snell et al., 1991). Regarding gender differences, studies show that men exhibit higher levels of sexual assertiveness (Snell et al., 1991).

1.6.8. Sexual optimism

Sexual optimism is defined as the expectation that the sexual aspects of one’s life will be positive and rewarding in the future (Snell et al., 2001). Men’s and women’s scores on sexual optimism has been found to be quite similar (Snell et al., 2001).

1.6.9. Sexual problem self-blame

Sexual problem self-blame is defined as the tendency to blame oneself when the sexual aspects of one’s life are unhealthy, negative or undesirable in nature (Snell et al., 2001). People with higher sexual problem self-blame attribute problems in the sexuality realm to themselves, resulting in accompanying negative emotions and cognitions. Self-blame is a component of self-directed emotions like guilt or depression. No evidence, however, has been found for links between sexual problem self-blame and other sexual self-concept aspects. Men have been found to have a higher tendency for sexual problem self-blame (Snell et al., 2001).

1.6.10. Sexual monitoring

Sexual monitoring is defined as the tendency to be aware of the public impression of one’s sexuality (Snell et al., 2001). This aspect measures an individual’s concern for how their sexuality is perceived. People with greater sexual monitoring tend to be more sensitive to others’ evaluations of their sexual history or sexual partners and concerned with how “sexual” others consider them to be (Snell et al., 1991; Snell & Wooldridge, 1998). Sexual monitoring has been found to negatively correlate with sexual anxiety among both genders and sexual satisfaction among women (Snell et al., 1991).

1.6.11. Sexual motivation

Sexual motivation is defined as the motivation and desire to be involved in a sexual relationship (Snell et al., 2001). It is therefore natural that people with greater sexual motivation report more

sexual partners (Snell et al., 2001). Consistent with the traditional stereotype about male sexuality, men have been repeatedly found to have greater sexual motivation (Kenney, Thadani, Ghaidarov, & LaBrie, 2013; Snell et al., 2001).

1.6.12. Sexual problem self-management

Sexual problem self-management is defined as the tendency to believe that one has the capacity and skills to effectively manage and handle any sexual problems that one might encounter (Snell et al., 2001). It has been found that women with more sexual experience were more likely to believe that they hold the necessary skills and abilities to handle any future sexual problems. However, they were also more willing to blame themselves if they encountered such problems in their future sexual relationships (Snell et al., 2001).

1.6.13. Sexual self-esteem

Sexual self-esteem is defined as the general tendency to positively evaluate one’s own capacity to engage in healthy sexual behavior and to experience one’s sexuality in a satisfying and enjoyable way (Snell et al., 2001). It is important in order to maintain interpersonal functioning as well as to develop and maintain a healthy sex life (Brassard, Dupuy, Bergeron, & Shaver, 2015). Studies report a positive relationship between sexual self-esteem and sexual satisfaction as well as sexual assertiveness, therefore, targeting sexual self-esteem can have a positive effect on both sexual satisfaction and sexual assertiveness (Menard & Offman, 2009; Oattes & Offman, 2007). Interestingly, sexual self-esteem is also positively correlated with sexual preoccupation (Snell et al., 1992). Studies show that women have lower sexual self-esteem scores compared to men (Brafford-Squiers, 1998).

1.6.14. Sexual satisfaction

Sexual satisfaction is defined as the tendency to feel positive affective responses arising from a subjective evaluation of the sexual aspects of oneself (Snell et al., 2001). Studies have repeatedly found a strong, positive association between sexual satisfaction and overall relationship satisfaction. Sexual satisfaction is also found to be strongly positively linked to sexual self-esteem and sexual consciousness (Snell et al., 1991; Menard & Offman, 2009). Despite the overall large body of research on sexual satisfaction, the results on how age, gender or other variables affect it

are mixed. Some studies have shown sexual satisfaction to decrease with age (Ojanlatva et al., 2003; Greeley, 1991), while others found this pattern only amongst women (Laumann et al., 1994) or only men (Cheung et al., 2008).

1.6.15. Power-other sexual control

Power-other sexual control is defined as the belief that the sexual aspects of life are controlled by others who are more powerful and influential than oneself (Snell et al., 2001). Together with control by chance/luck, control by powerful others are the two sources constituting external locus of control, as opposed to internal locus of control (Marshall, 1991). Power-other sexual control has been found to positively correlate with sexual monitoring and sexual depression, and negatively correlate with sexual self-esteem and sexual consciousness among women (Snell et al., 1991; Snell et al., 1992).

1.6.16. Sexual self-schema

Sexual self-schema is a cognitive generalization about the sexual-related aspects of oneself (Hoyt & Carpenter, 2015). It is derived from previous experiences and manifests itself in present experiences, thereby influencing sexually relevant information processing and guiding future sexual behavior (Andersen & Cyranowski, 1994). People with well-developed sexual self-schemas tend to be more willing to enter intimate relationships, and experience more positive emotions in sexual situations including higher levels of sexual arousal (Hoyt & Carpenter, 2015; Cyranowski & Andersen, 1998). Furthermore, positive sexual self-schema is also linked to higher sexual consciousness and safe sex self-efficacy (Horne & Zimmer-Gembeck, 2006). Sexual self-schema is among one of the better researched aspects of the sexual self. However, it is often used in parallel with sexual self-concept, which can lead to considerable confusion and difficulty to differentiate between them.

1.6.17. Sexual fear

Sexual fear is defined as the fear of engaging in sexual relations with another individual (Snell et al., 2001). Although not studied extensively, research relating to sexual fear has touched upon topics such as fear of intimacy and abandonment, sexual abuse, sex during pregnancy and sexual pain disorders. Women have reportedly greater sexual fear compared to men (Snell et al., 2001).

For men, sexual fear seems to negatively correlate with more sexual experiences (Snell et al., 1993). This is understandable: individuals with aversive affective reactions to sex, like greater sexual anxiety, fear or depression are naturally less inclined to engage in sexual relations (Snell et al., 1993).

1.6.18. Sexual problem prevention

Sexual problem prevention is defined as the belief that one has the ability to prevent oneself from developing sexual problems or disorders (Snell et al., 2001). Women with more sexual experience were more likely to perceive themselves as capable of preventing and handling any future sexual problems that might occur to them. Nonetheless, they were also more likely to blame themselves for any such future sexual problems (Snell et al., 2001).

1.6.19. Sexual depression

Sexual depression is defined as the chronic tendency to feel sad, disappointed, and depressed about one’s sex life and capability to relate sexually to another individual (Snell et al., 2001; Snell & Papini, 1989). Sexual depression negatively correlates with sexual experience (Snell et al., 1993), sexual satisfaction (Offman & Matheson, 2005), sexual self-esteem (Snell & Papini, 1989) and sexual consciousness (Snell et al., 1991). Amongst men, sexual depression was also positively correlated with sexual preoccupation (Snell & Papini, 1989), while for women, a positive correlation was found between sexual depression and sexual monitoring (Snell et al., 1991). A positive association has been found between sexual anxiety and sexual depression (Offman & Matheson, 2005).

1.6.20. Internal sexual control

Internal sexual control, or personal sexual control, is defined as the belief that the sexual aspects of one’s life are determined by one’s own personal control (Snell et al., 2001). Internal control is an overall important moderator of the linkage between age and health (Marshall, 1991). Greater knowledge of own personal resources, skills and capabilities counteracts any decline in personal infirmity that accompanies aging, also relating to sexual health.

1.7. Research questions and hypotheses

The aim of this study is to present and explore the structures of human sexuality by investigating the Multidimensional Sexual Self-Concept Questionnaire (MSSCQ) in a sample of 12,449 English-speaking self-selected respondents from a network perspective. The network theory suggests a novel way of conceptualizing psychological constructs as interacting systems that actively contribute to the construct (Briganti, Fried, & Linkowski, 2019). According to it, psychological variables such as symptoms or aspects of a particular construct directly affect each other instead of being caused by an underlying latent entity (Epskamp & Fried, 2018). In the past decade, a growing body of research has been viewing mental disorders as networks of interacting symptoms which arise due to direct interactions and causal dynamics between the symptoms (Fried et al., 2017). Variables in the network (e.g. aspects of a construct or symptoms of a disorder) are represented by “nodes” and relationships between these variables are represented by “edges” that connect the variables. Targeting the central symptoms within a network (i.e. centrally positioned nodes) should affect the entire network structure. Identifying these central symptoms might therefore offer possible targets for intervention (Fried et al., 2017). The network theory and methodology is being increasingly applied in psychopathology research and touch upon such topics as depression (Fried et al., 2016; Wasil et al., 2020), posttraumatic stress disorder (Spiller et al., 2017; Armour et al., 2017; Sullivan et al., 2018), obsessive-compulsive disorder (Cervin et al., 2020) and eating disorders (Olatunji et al., 2018) among many others. However, the application of the network perspective in sexual health research is less common and, hence, all the more valuable. A network analysis has been conducted on the issue of female sexual function (Gunst et al., 2018), hypersexuality (Werner, Stulhofer, Waldorp & Jurin, 2017), sexual health knowledge and practices (Valles et al., 2017), yet no prior studies have focused on sexual self-concept. Such conceptualization can be insightful, thus proposing a different point of view on this integral part of human sexuality and providing a better understanding on how its aspects are interconnected and the nature of relationships they hold. This can subsequently inform future sexual health research on possible approaches to strengthen sexual self-concept and contribute to sexual self-concept assessment innovation.

Three exploratory research questions were raised stemming from the aforementioned aim: 1. How connected are different aspects of sexual self-concept and which aspects are most central (i.e. interconnected)?

2. Does the sexual self-concept network differ between genders?

3. Does the sexual self-concept network differ between different age groups?

The first research question follows findings of various connections between the aspects of sexual self-concept and offers clinical implications, as knowledge on centrality and the interactions that take place within the network can enable new insights into the core and possibly influential aspects of sexual self-concept, thus offering a better understanding of sexual self-concept and contributing to its assessment innovation. The second research question stems from Snell, Jr Snell and editor’s (2001) observations about specific aspects of sexual self-concept differing between genders, such as motivation to avoid risky sex and sexual fear being more prevalent among females, males reporting greater sexual preoccupation, sexual problem self-blame, and chance/luck sexual control. The last research question stems from two areas of research findings. Firstly, the majority of researchers agree on sexual self-concept generally becoming more positive over time, secondly, however, several aspects of it (e.g. sexual satisfaction) have been found to not follow this general tendency of improvement. It would be therefore beneficial to examine whether, and if so- how, sexual self-concept network varies between different age groups.

Furthermore, based on corresponding research findings by various researchers (e.g. Anticevic et al., 2017; Menard & Offman, 2009; Oattes & Offman, 2007), it was hypothesized that the relation (edge) between sexual self-esteem and sexual satisfaction would be the strongest relation (edge) in the network.

2. Methods

2.1. Participants

The data used in the present study was obtained from the Open Source Psychometrics Project (https://openpsychometrics.org) that makes all of its data freely available online. The participants constituted a self-selected sample who found the test through the list of personality tests made available on the aforementioned website or through reference links from social media. All participants provided written informed consent. The dataset included only those respondents’ data, who had confirmed that their answers were accurate and suitable for research upon completion, and, due to the Child Online Protection Act, who specified their age to be above 13. The dataset consisted of responses to the MSSCQ (Snell, 1998) along with age and gender of the respondent.

The MSSCQ data was collected from 2012 to early 2014, and the full dataset (N = 17,685) was uploaded to the website on September 18th_{, 2014. Participants were included in the present analysis} if they completed all 100 MSSCQ questions, reported an age between 13 and 80 years and specified their gender. From the original 17,685 respondents that the dataset includes, 5,236 (29,61%) were removed due to: 1. missing values on the MSSCQ (n=4,896); 2. identifying with a gender other than male or female or not specifying any (n=331); 3. or not meeting the age of interest (n=9). This effectively reduces the final sample size to 12,449 respondents (70,39%). Overall, the data has a notably big advantage – a very large sample size, however, at the cost of having somewhat less data on other demographic variables (e.g. education, marital status, ethnicity), which therefore limits the implications and explanatory power one can draw from the results.

2.2. Measures

Sexual self-concept was assessed using The Multidimensional Sexual Self-Concept Questionnaire (MSSCQ; Snell, 1998). The MSSCQ is a validated self-report instrument, consisting of 100 items and designed to measure 20 psychological aspects of human sexuality: 1. sexual anxiety, 2. sexual self-efficacy, 3. sexual consciousness, 4. motivation to avoid risky sex, 5. chance/luck sexual control, 6. sexual preoccupation, 7. sexual assertiveness, 8. sexual optimism, 9.sexual problem self-blame, 10. sexual monitoring, 11. sexual motivation, 12. sexual problem management, 13. sexual esteem, 14. sexual satisfaction, 15. power-other sexual control, 16. sexual self-schemata, 17. sexual fear, 18. sexual problem prevention, 19. sexual depression, 20. internal sexual control. A 5-point Likert-type scale was used to evaluate the items, ranging from 1 to 5: not at all characteristic of me (1), slightly characteristic of me (2), somewhat characteristic of me (3), moderately characteristic of me (4), and very characteristic of me (5), with some reverse-scored items included (i.e. item 27, 47, 68, 77, 88 and 97). The items with reverse scores were inverted before further averaging and data analysis. The items on the MSSCQ subscales alternated in ascending numerical order for each subscale (i.e. Subscale 1 consisted of Item 1, 21, 41, 61, 81; Subscale 2 consisted of Item 2, 22, 42, 62, 82) (Fisher et al., 2011). The items on each subscale were averaged and ranged from 1 to 5. Higher scores corresponded to higher levels of each tendency, with a higher value meaning a “better”, “healthier” score in a tendency. The participants thus received 20 separate scores, one for each of the 20 MSSCQ subscales (Snell, 1995).

Furthermore, before performing further analysis, three aspects of the MSSCQ (sexual anxiety, sexual fear and sexual depression) were reverse scored for the sake of consistency, where the higher value corresponds to a “more positive” score in a tendency. This is based on the MSSCQ author and other studies distinguishing the 3 aspects as the negative dimensions of the scale (Snell et al., 2001; Ziaei et al., 2013). For instance, in a study by Ziaei et al. (2013), the researchers found the three aspects to correlate positively with each other and negatively with the rest of the MSSCQ aspects.

2.3. Research design and statistical methods

In this study, a novel perspective in sexual health research was employed—a network analysis. The network structure of sexual self-concept consisted of sum scores of the MSSCQ aspects that are represented as interconnected nodes. Between-person networks are commonly estimated using regularized partial correlations (Epskamp & Fried, 2018). The Gaussian Graphical Model (GGM; Lauritzen, 1996), a regularized partial correlation network was estimated. This model is the most popular and commonly used method to estimate psychological networks (Epskamp & Fried, 2018). When visualizing the GGM as a graph, nodes depict psychological variables, and edges between them represent unknown statistical relationships, calculated as regularized partial correlations (Epskamp & Fried, 2018). The edges are shared variances between the variables. More precisely, an edge between two nodes indicates that these two variables are related after controlling for their associations with all other nodes in the network (Epskamp & Fried, 2018).

Data analysis was performed with the open source statistics program JASP (version 0.12.2). It automates many steps that would have to be performed by code when estimating network models in R. The analysis was conducted in three steps: network estimation, network inference and network comparison (Fried et al., 2018). EBICglasso (Extended Bayesian Information Criterion Graphical Least Absolute Shrinkage and Selection Operator) was chosen as the regularized estimation method. The EBICglasso estimates partial correlations between all variables, shrinking the absolute weights to zero and thus eliminating spurious edges. A general network, two separate gender-based networks, and four age group-based networks were estimated. Two parameters are important in regard to lasso regularized network estimation: tuning parameter (lambda) and hyperparameter (gamma). Tuning parameter lambda controls the level of sparsity in order to retrieve the true network structure, while hyperparameter gamma controls the strictness of the

network regularization (Epskamp & Fried, 2018). JASP automatically sets the EBIC hyperparameter gamma to γ = 0.5, and the lasso tuning parameter lambda to Λ=0.5. Moreover, JASP also automatically implements the Fruchterman-Reingold algorithm (Fruchterman & Reingold, 1991) for positioning the nodes, which organizes the network according to the strength of connections between the nodes. The correlation method was set on automatic, which is namely the polychoric correlations method implemented for ordinal variables. Network inference was computed by exploring node centrality, inspecting three centrality measures: betweenness, closeness and strength. Betweenness centrality measures how often a certain node lies on the shortest path between every other pair of nodes, indicating how important a node is in the average path between two other nodes (Costantini et al., 2015). Closeness centrality measures the average distance from the node to all other nodes in the network, indicating how well a node is indirectly connected to other nodes (Costantini et al., 2015). Strength centrality, also known as degree centrality, is the sum of a node’s connection weights, indicating how well a node is directly connected to other nodes (Costantini et al., 2015). Therefore, betweenness indicates the extent of facilitating the flow of information in a network, closeness quantifies how fast a node can be reached via other nodes, and strength centrality suggests how directly impactful (or impacted) a node is (Costantini et al., 2015; Opsahl, Agneessens, & Skvoretz, 2010). Finally, network comparison was done by comparing edge weights and centrality measures. The network plot and centrality plot were used to graph the results. Furthermore, the colorblind node pallet was utilized, where blue edges between nodes indicate positive partial correlations, and red edges indicate negative partial correlations. Wider and more saturated edges portray stronger partial correlations (Epskamp & Fried, 2018). Variable names were shown in a legend due to their extensive length.

To address the first research question, namely the connectivity of the network, centrality measures and edge weights were analyzed. To assess the second research question, whether networks vary between men and women, gender was included as the covariate to split the results by. Edge weights, centrality measures as well as overall networks were compared. Furthermore, networks were estimated for different age groups, thus addressing the last research question. For age group comparison, the sample was allocated to four different age groups: adolescents (13-18 years), youth (19-30 years), adult (31-45 years), and older adult and seniors (46+ years) as per the Canadian Federation of Sexual Health (2006) classification. In the used classification, the older

adults and seniors are two separate age groups, however, due to an insufficient amount of respondents in these age groups to perform a network analysis, they were merged into one.

3. Results

3.1. Demographic characteristics

The 12,449 respondents included in the final sample were on average 26.31 years old (SD=10.74) and a slight majority of the sample (57,39%) was female. Men respondents were on average older (M=28.33, SD=11.91) than women (M=24.8, SD=9.51). No other demographic information was available from this sample. Descriptive statistics of the respondents age can be found in Table 1 in Appendix 1.

3.2. Network analysis of the sexual self-concept

Figure 1 illustrates the estimated network of sexual self-concept. It features no unconnected nodes in the network, and 160 of all possible 190 edges (84%) were estimated to be above zero (61 negative and 99 positive edges). The strongest (i.e. positive) edges in the network are between 1. sexual satisfaction and sexual depression, 2. sexual anxiety and sexual depression, 3. sexual problem self-blame and sexual problem management, 4. sexual preoccupation and sexual motivation, and 5. sexual anxiety and sexual fear, with respective edge weights between 0,352 and 0,429. A number of negative edges were obtained, among which the strongest are between 1. power-other sexual control and internal sexual control, and 2. sexual anxiety and sexual monitoring, with both edge weights slightly over -0,200. Motivation to avoid risky sex and sexual depression contained the highest number of negative edges (n=10) while sexual esteem had the most positive edges (n=14). Overall, motivation to avoid risky sex and sexual optimism held the highest amount of edges (n=18) in absolute terms: respectively 10 negative and 8 positive, and 12 positive and 6 negative edges. Sexual self-assertiveness and sexual satisfaction held the least (n=14), respectively 10 positive and 4 negative, and 9 positive and 5 negative edges.

Centrality estimates are presented in Figure 2 and Table 2 (the highest scores in each measure bolded). As seen from both the graph and table, sexual motivation is the node with the highest strength centrality in the network, while sexual anxiety and sexual depression score the highest on betweenness and closeness centrality measures.

Figure 1. Sexual self-concept network structure.

Table 2. Centrality scores per variable

Network

Variable Betweenness Closeness Strength

V1. Sex. anxiety (r) 2.028 1.767 1.270 V2. Sex. Self-Efficacy -0.461 0.519 -0.859 V3. Sex. Consciousness -0.315 0.014 0.619 V4. Motivation to avoid risky sex -0.608 -1.168 -1.233 V5. Chance/luck sexual control -1.193 -0.669 -1.735 V6. Sexual Preoccupation -0.022 0.293 0.210 V7. Sex. Self-Assertiveness -1.632 -0.935 -1.155

V8. Sex. Optimism -0.022 0.348 -0.334

V9. Sexual Problem Self-Blame -1.047 -1.426 0.368 V10. Sex. Monitoring -0.461 -0.100 -1.505 V11. Sex. Motivation 0.417 0.406 2.006

V12. Sex. Problem Management 0.124 -1.028 0.176

V13. Sex. Esteem -0.461 0.463 1.362

V14. Sex. Satisfaction 0.564 1.447 0.192 V15. Power-Other Sexual Control -0.168 -0.605 -0.244 V16. Sex. Self-Schema -1.047 -0.821 -0.718

V17. Sex. Fear (r) 0.710 0.752 0.396

V18. Sex. Problem Prevention 1.003 -0.545 0.145 V19. Sex. Depression (r) 2.320 2.130 1.068 V20. Internal Sex. Control 0.271 -0.841 -0.029

3.3. Network structure of sexual self-concept across gender

Men and women’s sexual self-concept networks are presented in Figure 3. For visual comparison, two separate networks were determined. In the men’s network, 153 edges (80,5%) were estimated to be above zero, and in the women’s network 160 (84%) were estimated to be above zero. A number of strong edges were obtained, among which the strongest in the men’s network were 1. sexual anxiety and sexual depression, 2. sexual problem self-blame and sexual problem management, 3. sexual depression and sexual satisfaction, 4. sexual motivation and sexual preoccupation, 5. sexual esteem and sexual satisfaction and lastly 6. sexual problem management and internal sexual control, with respective edge weights between 0,330 and 0,434. In the women’s network, the strongest edges were between 1. sexual satisfaction and sexual depression,

2. sexual anxiety and sexual depression, 3. sexual problem self-blame and sexual problem management, 4. sexual anxiety and sexual fear, 5. sexual preoccupation and sexual motivation and lastly, 6. sexual motivation and sexual fear, with respective edge weights between 0,359 and 0,444. The strongest negative edges were the same for both genders: between sexual anxiety and sexual monitoring, and power-other sexual control and internal sexual control, with both edge weights slightly over -0,200.

Figure 3. Gender-based sexual self-concept network structures. Men’s network (left) and women’s network (right).

Centrality measures of the gender-based networks are presented in Figure 4. In both networks, sexual motivation was found to be the strongest node in the network, while sexual anxiety and sexual depression scored the highest on betweenness and closeness centrality measures. Centrality scores can be found in Table 3 in Appendix 2.

3.4. Network structure of sexual self-concept across age groups

The four age groups differed in size and gender proportion. The largest age group was the youth group (n=6,380) and comprised over half (51,25%) of the total sample size. The youth group had slightly more women (58,84%) than men. The second biggest age group was the adolescent group (n=2,947), where women (66,41%) were the majority. The adult group had an almost equal number of men and women, whereas the older adult and senior age group consisted of more men (61,06%) than women. The corresponding networks and centrality measures of each age group are presented in Figure 5-8. Figure 9 provides comparison of centrality measures between different age groups.

Figure 5. Adolescent group (13-18 years) sexual self-concept network structure and centrality measures

Figure 6. Youth group (19-30 years) sexual self-concept network structure and centrality measures

Figure 7. Adult group (31-45 years) sexual self-concept network structure and centrality measures

Figure 8. Older adult and senior group (46-78 years) sexual self-concept network structure and centrality measures

4. Discussion

The aim of the study was to investigate the structure of sexual self-concept in a large sample of 12,449 individuals using network analysis. Below, the core research questions and results are briefly reiterated, and subsequently analyzed in the context of the existing literature. The research questions sought to examine how the aspects of sexual self-concept are situated in a network and whether the structure of the networks differs between genders and different age groups. The estimated networks were undirected weighted networks, meaning that all relations in the network were symmetrical and had specific weights – connection strengths. The estimated networks featured no unconnected nodes and consisted of both positive and negative edges, and the strongest edges as well as centrality measures were mostly shared among both genders and different age groups.

Figure 9. Centrality measures of sexual self-concept network nodes in different age groups

4.1. Sexual self-concept connectedness and centrality

Overall, the network of sexual self-concept was found to be strongly interconnected with both positive and negative partial correlations. According to Robinaugh, Millner and McNally (2016), negative edges are typically more prevalent in non-clinical populations, which is the case for the current sample. However, this might be due to the used questionnaire consisting of both positive and negative aspects of sexual self-concept. Although sexual anxiety, sexual fear and sexual depression were reverse scored due to their negative nature (meaning higher scores meant lower tendency on these aspects), such aspects as sexual preoccupation or sexual problem self-blame also carry a negative connotation to them, thus making the boundary between positive and negative aspects of sexual self-concept unclear. Nevertheless, the measurement of sexual self-concept is not based on one total score, but rather individual scores of each aspect. Therefore, attention should be directed towards the individual aspects of sexual self-concept and the interactions between them, which the network approach can assist in.

To examine the connectivity of the network and thus address the first research question, centrality measures and edge weights were analyzed. A number of findings deserve a closer look, namely the strongest edges found in the general sexual self-concept network (Figure 1). Sexual depression and sexual satisfaction share the strongest connection in the network, indicating a link between feeling depressed about your sex life and feeling less sexual fulfilment (and vice versa). Previous literature is consistent with these findings, noting that such factors as sexual depression can contribute to lowered sexual satisfaction. Yet researchers are careful to note that sexual depression is not equivalent to a lack of sexual satisfaction (Offman & Matheson, 2005). The same kind of connection is found between sexual anxiety and sexual depression, meaning that in regards to how people feel about themselves as sexual beings, feeling worried and anxious is closely related to feeling sad. This is in line with previous studies that report a positive association between these two negative aspects of sexual self-concept (Offman & Matheson, 2005). Furthermore, an interesting finding shows the tendency to blame oneself for undesirable aspects of one’s sexual life to be connected to the belief that one is equipped with the skills to effectively manage arising sexual problems. This means that people who perceive themselves as capable of managing problems in their sex lives also tend to blame themselves if they encounter problems in this area of their lives. Previous research has found similar results, although only among women (Snell et al., 2001). What is more, a strong edge between sexual preoccupation and sexual motivation

suggests that people who excessively think about sex also tend to long for sexual interactions, and vice versa. Previous research has also found such a link (Snell et al., 1991). Another strong edge that emerged from the network is between sexual anxiety and sexual fear. This indicates that feeling apprehensive about engaging in sexual relations is closely connected to feeling uncomfortable about your sexual life and one might reinforce the other. This is supported by previous research on sexual anxiety, that also found it associated to stronger abstinence attitudes, lower perceived sexual readiness, fewer reports of having a partner and less sexual experience (Hensel et al., 2011). Interestingly, despite strong edges between sexual anxiety and sexual depression, and sexual anxiety and sexual fear, sexual fear and sexual depression does not share such a connection, contrary to previous findings that associate the three negative aspects of sexual self-concept together (Snell et al., 1993). Lastly, touching upon the hypothesized edge between sexual self-esteem and sexual satisfaction, although not the strongest edge as expected, it is among the strongest connections in the men’s sexual self-concept network and a nonetheless strong connection in the network.

A few relatively strong negative edges also emerged in the network. The observed strong negative edge between sexual anxiety and sexual monitoring suggests that being concerned about how your sexuality is perceived is connected to feeling tense and worried about the sexual aspects in your life and vice versa. The strong negative connection between power-other and internal sexual control is understandable due to the opposing meanings of the aspects, former localizing the control of one’s sexual life externally, and the latter – internally.

The network was further analyzed by investigating the centrality of the sexual self-concept aspects. Centrality measures represent the connectedness of an aspect to other aspects and quantify their positions within the network (Robinaugh et al., 2016). This is commonly assessed with 3 indices: strength, betweenness and closeness. Results show sexual motivation as the node with the highest strength centrality measure in the network, meaning it could be assumed that sexual motivation directly interacts or is associated with many other aspects in the sexual self-concept network. Furthermore, results indicate that sexual depression and sexual anxiety have the shortest average distance to the rest of the network, with sexual depression slightly higher in the measure. This could possibly mean that the information or stimuli from sexual depression and sexual anxiety reaches other nodes the quickest and changes in these two aspects are more likely to quickly affect the rest of the sexual self-concept network, and vice versa. However, it is interesting that these two