Heterosexual Script in Italian Young Adults: Measurement Across Genders

The Heterosexual script has not yet received extensive scientific attention in, one of the countries in which gender stereotypes are particularly prominent. The purpose of this study was to examine adherence to sexual scripts among Italian young adults by first evaluating the dimensionality, psychometric properties, and gender invariance of the Italian version of the Heterosexual Script Scale (HSS), a 22-item self-report questionnaire measuring endorsement of sexual scripts in the context of heterosexual relationships. A community sample of 2781 women and men (F = 50.5%) was recruited in 2021 (mean age = 22.32, SD = 3.53 years, age range: 18–35 years). Participants completed the Italian version of the HSS, measures of ambivalent sexism, and a socio-demographic questionnaire. Exploratory factor analysis revealed that, after deletion of four items (items 4, 7, 16, 22), four factors should be extracted: Courtship Strategies, Women Depend on Men, Women’s Self-Sacrifice, and Sex Defines Masculinity; Women Set Sexual Limits. The confirmatory factor analysis indicated that the 18-item four-factor structure had a good fit (RMSEA [90% CI] = .06 [.06; .07]; CFI = .95; SRMR = .05). Good internal consistency (Cronbach’s alphas for the four factors ranged from .65 to .72), construct validity, and metric invariance across genders were demonstrated. Overall, findings indicate that the HSS can be used to reliably assess the endorsement of gendered sexual scripts among young Italian women and men. Some specific cultural aspects emerged. Researchers and sexual health practitioners should consider context-specific sexual scripts when designing research and intervention protocols. Sex education curricula could be improved by integrating modules aimed at promoting egalitarian gender scripts.

The Sexual Script Theory posits that virtually all human behaviors, including sexuality, are subjected to some sort of cultural scripting (Gagnon & Simon, 1973, 1987Simon & Gagnon, 1986). Sexual scripts are sets of culturally shared norms that inform individuals on how to respond to sexual clues and define the appropriate sexual conduct (Simon & Gagnon, 1986). In most cultures, sexual scripts are gendered, meaning that men and women are assigned to different sexual roles (Kim et al., 2007;Sakaluk et al., 2013;Seabrook et al., 2016;Simon & Gagnon, 1986;Wiederman, 2005).
In Westerner societies, the only sexual script that is proposed as normative is the one of heterosexual encounters, and it consists of a quite rigid set of norms that prescribe opposing, but complementary, roles for men and women (Kim et al., 2007). Specifically, the "Heterosexual Script" (Kim et al., 2007) is composed of three core elements: the sexual double standard, gender-specific courtship strategies, and gender-specific orientations toward commitment (Kim et al., 2007). The sexual double standard refers to the common belief that men are sexual pursuers and are always willing to engage in sex, while women are expected to set sexual limits and behave as "gate-keepers". Gender-specific courtship strategies define the appropriate courtship techniques for each gender. Men are expected to attract women with power and resources and to be active initiators. Women, instead, are expected to take on a passive role and wait to be chosen by men. Finally, gender-specific orientations toward commitment refer to the idea that men tend to avoid commitment, while women attribute a great value to the romantic relationship and are prone to make personal sacrifices to maintain the relationship (Kim et al., 2007).
The norms of the Heterosexual Script reflect underlying sexist attitudes. Sexism can be broadly defined as a form of prejudice and discrimination based on a person's gender (Glick & Fiske, 1996. Sexism encompasses two sets of ambivalent beliefs: hostile and benevolent sexist attitudes. Hostile sexism consists of overtly hostile and resentful beliefs (e.g., "women seek to control men through sexuality" and "men are arrogant") (Glick & Fiske, 1996, 2001. On the other hand, benevolent sexism consists of a set of subjectively positive attitudes (e.g., "women need to be protected by men" and "men need to be taken care of") which usually elicit prosocial behaviors, but stem from the belief that the other gender is inferior (Glick & Fiske, 1996. Traditional sexual scripts are learned during childhood through experiences of socialization with family members and peers (Simon & Gagnon, 1986). Researchers in the fields of social and developmental psychology have proposed different theories to explain the process of gender construction and differentiation during childhood. Social learning theories propose that young children develop beliefs regarding what is to be considered appropriate behavior for men and women by observing their environment and, in particular, the significant adults they identify with. Moreover, the internalization of gender roles can take place through a process of vicarious learning, in which children observe and assess how others respond and react to the different behaviors of their "gender models" (Endendijk et al., 2018). Cognitive theories of gender constructivism posit that when children learn that two distinct genders exist, they develop a gender scheme. This gender scheme is then used to process information based on this categorization. In this way, children start perceiving their environment in terms of gender and, ultimately, they adapt their behaviors and construct their identity based on the expectations they have about their gender (Bem, 1981). The internalization of gender schemes is further supported by the different socialization paths to which young boys and girls are exposed. In this regard, solid evidence shows that children internalize and apply gender stereotypes from a very young age (King et al., 2021;Solbes-Canales et al., 2020). Gender stereotypes are applied to numerous life domains, from the choice of toys and activities to the construction of one's personality characteristics and career ambitions (King et al., 2021;Solbes-Canales et al., 2020). Ultimately, gender stereotypes constitute the underlying foundation of gender discrimination (Burgess & Borgida, 1999).
Growing up, sexual scripts are further internalized through repeated exposure to the media. Indeed, many studies have documented the overwhelming presence of gendered scripts in TV programs (Kim et al., 2007), social media (Rodriguez- Sanchez et al., 2020), magazines (Hust et al., 2016), and newspapers (Tartaglia & Rollero, 2015). Sexual scripts become particularly important during emerging adulthood when most individuals experiment with romantic and sexual relationships (Seabrook et al., 2016). In this regard, numerous studies have shown that sexual scripts are prevalent on college campuses, as college students tend to refer to traditional sexual scripts to evaluate, and respond to, sexual and romantic situations (e.g., Allison, 2019;Ford, 2020;Jozkowski et al., 2017). Likewise, traditional gender and sexual scripts have also been found to shape users' interactions in dating apps such as Tinder (Christensen, 2020;Comunello et al., 2021).
Endorsing the norms of the Heterosexual Script is associated with several negative sexual-related outcomes. In women, it is related to less sexual risk knowledge, protective behaviors, sexual satisfaction, sexual subjectivity, sexual assertiveness, and body self-esteem during sexual activities (Curtin et al., 2011;Sanchez et al., 2011;Scappini & Fioravanti, 2022). In men, greater adherence to the traditional masculine role is associated with a higher risk of perpetrating sexual violence against women (Check & Malamuth, 1983;Murnen et al., 2002) and with more negative attitudes toward gay men (Davies et al., 2012). Moreover, the pressure to comply with gender-specific expectations can lead both women and men to "go along with" sex and engage in sexual activities that they do not desire, to avoid shame or embarrassment (Ford, 2017(Ford, , 2020. In both genders, holding gender stereotypical attitudes is associated with increased rape myth acceptance and victim blame (Acock & Ireland, 1983;Davies et al., 2012;Willis, 1992). Finally, the heterosexual script was found to influence the first-date scripts of young adults (Cameron & Curry, 2019), as well as the patterns of sexual initiation and the sexual activities of young heterosexual couples (Vannier & O'Sullivan, 2010, 2012. Indeed, it was observed that, among young heterosexual people, male sexual initiation is significantly more common than female initiation (Vannier & O'Sullivan, 2010), which seems to reflect the traditional script of men as sexual pursuers and women as sexual gatekeepers. Moreover, it was found that women perform oral sex to their partners significantly more often than men, which seems to reflect the script according to which women have to please the sexual requests of their partners to keep them close and committed to the relationship (Vannier & O'Sullivan, 2012).

Measuring the Heterosexual Script
In light of the literature discussed so far, it appears fundamental to monitor the prevalence of gendered sexual scripts among young adults, to be able to inform relevant authorities about the need to implement prevention 1 3 programs aimed at reducing the impact that sexual scripts might have on young people's relational, sexual, and general well-being. In this regard, it seems particularly important to ascertain that the measures used to assess sexual scripts' endorsement have sound psychometric properties. Specifically, given the gendered nature of sexual scripts, assessing the gender invariance of such measures (i.e., assessing whether men and women tend to interpret and respond equally to these measures) seems to be particularly warranted.
In the last 50 years, several scales that measure the level of endorsement of traditional gender roles have been developed. However, as noted by Seabrook et al. (2016), none of these measures captures the complementarity of men's and women's roles within the specific context of heterosexual relationships. Most of the available gender roles measures have separate scales for men and women (see, for instance, the widely used Bem Sex Role Inventory, Bem, 1974; the Conformity to Masculine Norms Inventory, Mahalik et al., 2003;and the Conformity to Feminine Norms Inventory, Mahalik et al., 2005), which reflects an erroneous conceptualization of gender roles as distinct and independent, rather than interdependent (i.e., men are sexual pursuers and women are sexual gatekeepers). Another important limitation of several extant sexual scripts' measures is the fact that they are dated (e.g., the Sexual Script Questionnaire, LaPlante et al., 1980; the Attitudes Toward Women Scale, Spence & Helmreich, 1972;the Bem Sex Role Inventory, Bem, 1974). As sexual scripts evolve alongside cultural changes, older scales are unlikely to be able to reliably capture the level of gender scripts' endorsement in nowadays society. Furthermore, most of the existing measures are too narrow in focus, as they only assess some aspects of the heterosexual script. The Sexual Script Scale (Sakaluk, et al., 2013), for instance, focuses extensively on the sexual double standard, but it fails to capture relevant gender scripts regarding commitment (i.e., women make sacrifices to maintain the relationship vs men avoid emotional commitment) and courtship strategies (men attract women with power and resources vs women attract men with their bodies and looks). Finally, numerous scales measure gender roles' endorsement in domains and contexts other than romantic relationships, such as at work or school, or concerning family and domestic chores (e.g., the Femininity Ideology Scale, Levant et al., 2007;and the Attitudes Toward Women Scale for Adolescents, Galambos et al., 1985).

The Heterosexual Script Scale
In the attempt to overcome the aforementioned limitations of previous measures of sexual scripts' endorsement, Seabrook et al. (2016) developed a self-report questionnaire to assess the level of endorsement of the Heterosexual Script, following the framework provided by Kim et al. (2007): the Heterosexual Script Scale (HSS). The scale assesses the endorsement of traditional gender roles, elements of sexualization and objectification, and beliefs about romantic relationships in the specific context of heterosexual relationships. The validity of the HSS was tested on a mixed-gender sample (N = 555; 54.8% females) of university students from North America aged 17 to 26 (Seabrook et al., 2016). Exploratory factor analyses revealed that a 22-item four-factor solution showed the best fit to the data, which was then confirmed through a confirmatory factor analysis (Seabrook et al., 2016). The four HSS factors are distinct but related factors that load on a higher-order factor, the Heterosexual Script. Factor 1, Courtship and Commitment, assesses beliefs about committing to a heterosexual relationship. A sample item is "A woman should be willing to make personal sacrifices to satisfy her partner". Factor 2, Men as Powerful Initiators, assesses the belief that men actively look for sex while women passively wait for it. A sample item is "Men should be the ones to ask women out and initiate physical contact". Factor 3, Men Value Women's Appearance, assesses the belief that men are mostly interested in women's physical appearance. A sample item is "Being with an attractive partner gives a guy prestige". Finally, Factor 4, Sex Defines Masculinity/ Women Set Sexual Limits, investigates the idea that men are always ready for sex while women are responsible for limiting sexual intercourse. A sample item is "It is up to women to keep things from moving too fast sexually" (Seabrook et al., 2016). Respondents are asked to express their level of agreement with each of the 22 statements using a 6-point Likert-type scale that ranges from Strongly Disagree (1) to Strongly Agree (6). Higher scores indicate a stronger endorsement of the heterosexual script (Seabrook et al., 2016). The total scale showed good internal reliability (Cronbach's alpha = .88) and the subscales showed acceptable internal reliabilities (Cronbach's alphas for the four subscales were .76, .65, .74, and .67) (Seabrook et al., 2016). The convergent validity of the HSS was demonstrated by significant correlations with measures of endorsement of traditional gender roles, self-objectification, sexism, and self-sexualization (Seabrook et al., 2016). Concerning gender invariance, Factor 2 and Factor 4 of the HSS demonstrated full metric invariance across genders. However, two items from Factor 1 (items 2 and 4) and two items from Factor 3 (items 14 and 15) showed different loadings for men and women, which indicates that these particular items of the HSS may be interpreted differently by men and women. To the best of our knowledge, no other study has been conducted to test the psychometric properties of the HSS, especially gender invariance, outside of North America. Thus, it remains unknown whether the HSS can be reliably used in different cultural contexts. Moreover, validity evidence is missing considering separately males and females.

The Heterosexual Script in the Italian Context
Italy is one of the countries in which gender stereotypes are particularly prominent (Ragnedda & Budd, 2015), and therefore, they are likely to differentially shape the romantic and sexual experiences of Italian women and men. For a brief overview of the most common gendered beliefs in the Italian context, we hereby report the results of a national survey that was recently conducted on a large sample of Italian adults (N = 15034; age range 18-74 years) (ISTAT, 2019). The survey revealed that 32.5% of the respondents believed that being successful at work is more important for men than it is for women, and 27.9% of the sample believed that men should be responsible for the financial needs of the family (ISTAT, 2019). These results indicate that women in Italy are still expected to take on a rather passive role, while men are expected to be the providers and take care of their partners. Furthermore, 7.2% of the respondents believed that when women receive sexual advances they often refuse, but they actually mean to accept, and up to 23.9% of the respondents believed that women could provoke sexual aggression with their outfits (ISTAT, 2019). These results suggest that, on the one hand, Italian women are not yet fully accorded the right to assert their sexual choices and, on the other hand, that male sexual desire is still seen, by a large portion of the Italian population, as an urge that cannot be controlled. Even though recent evidence has shown that the sexual attitudes of young Italian adults are progressively shifting toward a more egalitarian and permissive direction (Minello et al., 2020), men and women in Italy are still being evaluated differently for their sexual behaviors (Migheli & Pronzato, 2020). In particular, Migheli and Pronzato (2020) found that, especially in the more conservative regions of Italy (i.e., the South), the heterosexual double standard is still applied to behaviors such as pre-marital sex, having multiple sexual partners, or having sex at a young age.
Several factors are contributing to the maintenance of gendered sexual scripts and gender stereotypes in the Italian context. Firstly, women in the Italian media are more often sexualized than those in other European countries (Ragnedda & Budd, 2015;Tartaglia & Rollero, 2015), which certainly contributes to reinforcing gender stereotypes and might also increase women's objectification (Fredrickson & Roberts, 1997). Furthermore, the Catholic religion has a strong impact on the socialization of Italian youth (Caltabiano et al., 2006), and Italians deem religion more important than the majority of other European citizens (De Santis et al., 2015). Given that the Catholic Church tends to promote traditional gender roles and sexual conservatism, it is possible that being religious might prompt a stronger adherence to the heterosexual script. Indeed, evidence shows that holding religious attitudes is associated with stronger adherence to traditional gender roles (Morgan, 1987); higher levels of sexism (Glick & Fiske, 1997;Maltby et al., 2010;Mikołajczak & Pietrzak, 2014); and a stronger endorsement of the sexual double standard (Emmerink et al., 2015;Migheli & Pronzato, 2020).

The Current Study
Following these premises, the present study aimed to examine adherence to sexual scripts among Italian young adults by first determining whether the HSS could be a psychometrically sound measure in the Italian context. To the best of our knowledge, the only available measures in the Italian language that can provide some insights on the adherence to gender stereotypes in the Italian context are the Ambivalent Sexism Inventory (Glick & Fiske, 1996; Italian version by Rollero et al., 2014) and the Ambivalence Toward Men Inventory (Glick & Fiske, 1999; Italian version by Rollero et al., 2014). However, those measures focus exclusively on sexism; therefore, they do not allow to capture the level of adherence to traditional sexual scripts in the specific context of heterosexual relationships.
For these reasons, the present study evaluated the dimensionality, internal consistency, gender invariance, and convergent validity of the Italian version of the HSS. This is important because it provides further cross-cultural research on the HSS, while also providing a useful tool with which to assess adherence to sexual scripts among Italian young adults. Based on the findings reported by Seabrook et al. (2016), we expected to find evidence for a four-factor structure for the Italian HSS, and we expected HSS scores to correlate positively with sexism scores.

Participants
A total of 2781 Italian women and men were recruited from the general population (%F = 50.5). The participants' ages ranged from 18 to 35 years (M = 22.32 years, SD = 3.53), 7.9% of the participants had an elementary or middle school diploma, 67.3% had a high school diploma, 22.55% had an academic degree, and 2.3% had postgraduate degrees. For what concerns religious faith, 29.9% of the participants stated that they were religious, while the remaining 70.1% stated that they were either atheists or agnostics.
The research protocol was advertised on social media platforms such as Facebook and Instagram. In particular, the research was advertised on a popular Instagram page that publishes daily content related to sexuality (Sessuologia, n.d.). Inclusion criteria were being between 18 and 35 years 1 3 of age and having or having had a heterosexual relationship in the six months preceding recruitment. Participation was voluntary and participants did not receive remuneration.
Given the recruitment procedure used for the current study, it is important to note that the present study employed a convenience sample, which cannot be considered statistically representative of the target population. In terms of education, for example, the current sample appeared to be slightly more educated than the general population in the same age group (i.e., in the current sample, fewer people had middle school as their highest level of education, and more people had an academic degree) (ISTAT, 2022).

Heterosexual Script Scale
The Italian version of the Heterosexual Script Scale (HSS, Seabrook et al., 2016) was obtained through a standard back-translation technique (Brislin, 1986). First, the parent English version was translated from English to Italian by a bilingual individual who was unaffiliated with the study (thus creating Version 1); next, the scale was translated back to English by a second individual (thus creating Version 2). Finally, Version 1 and Version 2 were compared by two additional independent translators who were not affiliated with the study. Minor discrepancies were settled through consensus. The Italian version of the HSS can be found in the Appendix.

Ambivalent Sexism Inventory and Ambivalence Toward Men Inventory
To test the convergent validity of the HSS, we followed the work of its original authors, and we administered the short versions of the Ambivalent Sexism Inventory (ASI; Glick & Fiske, 1996) and Ambivalence Toward Men Inventory (AMI; Glick & Fiske, 1999). The short versions of the ASI and AMI were created to enable a rapid assessment of sexism toward women and men respectively; they count 12 items each, and are both divided into two subscales, one assessing benevolent sexism and one hostile sexism. Participants are asked to express their agreement with each statement using a 5-point Likert scale ranging from 1 (Strongly Disagree) to 5 (Strongly Agree). The Benevolent Sexism subscale (BS) of the ASI includes items such as "Many women have a quality of purity that few men possess", and the Hostile Sexism subscale (HS) of the ASI includes items such as "Many women get a kick out of teasing men by seeming sexually available and then refusing male advances". The Benevolence Toward Men subscale (BM) of the AMI includes items such as "Men are more willing to put themselves in danger to protect others", and the Hostility Toward Men subscale (HM) of the AMI includes items such as "Most men sexually harass women, even if only in subtle ways, once they are in a position of power over them". Higher scores indicate that the respondent holds stronger sexist attitudes. The Italian short versions of both scales were found to have good psychometric properties and a level of reliability comparable to the original scales (Cronbach's alphas were .80, .85, .81, and .79 for the BS, HS, BM, and HM subscales respectively) (Rollero et al., 2014). In the current study, Cronbach's alphas were .76, .82, .79, and .70 for the BS, HS, BM, and HM subscales respectively.

Procedure and Statistical Analysis
The individuals who were interested in undertaking the survey were provided with a link redirecting them directly to the online research questionnaire. Information was given about the scope of the research, data treatment, and privacy. The anonymity of the participants was fully guaranteed, as participants were not asked to reveal any personal information that could be linked to their identity. Before starting the survey, participants were asked to declare that they were at least 18 years old and that they had read and understood the informed consent. International ethical guidelines were followed, and ethical approval for the study was received from the Institutional Review Board of the University of Florence.
To verify data reliability we checked for validity, completeness, and uniqueness. In particular, we verified that the information in our dataset included (i) only values in the required response range, (ii) no missing data, and (iii) no duplicated information. To examine the psychometric properties of the Italian HSS, a two-step analytic strategy was used. First, we split the dataset so that the first subsample had a random allocation of 1394 participants (702 women and 692 men) and the second sub-sample included 1387 participants (702 women and 685 men). The sample was randomly split using a computer-generated random seed. There were no significant differences between the two subsamples in terms of age, or in the distribution of all the other study variables. In the first subsample, we assessed the factor structure of the Italian HSS. The sample sizes in this subsample met a conservative item-to-participant ratio of 10:1 (Tabachnick & Fidell, 2013). Using the original validation study (Seabrook et al., 2016) as a guide, we used principal axis factoring with Oblimin rotation. The number of factors to be extracted was determined by factor eigenvalues above 1.0 (the EGV1 criterion) and examination of the scree plot, which are adequate criteria when a single factor is expected (Preacher & MacCallum, 2003). We then examined internal consistency by computing Cronbach's reliability coefficient. Next, data from the second subsample were subjected to CFA with maximum likelihood (ML) estimation method by using LISREL. Hypothesized modelling was based on the results of the EFA in the first subsample. Standard goodness-offit indices were selected a priori to assess the measurement models (Hu & Bentler, 1999): the normed model chi-square (χ 2 /df), the Comparative Fit Index (CFI), the standardized root mean square residual (SRMR), and the root mean square error of approximation (RMSEA). Specifically, a χ 2 /df value of < 5.00 indicates a good fit; CFI values should be close to or > .95 for a good fit but can be as low as .90 for an adequate fit. A cutoff value for SRMR is recommended to be close to or < .09. RMSEA values close to .06 indicate a good fit, with values ranging to .10 representing a mediocre fit. Gender invariance analyses were conducted by performing hierarchically nested confirmatory factor analyses (see Byrne, 2004, for testing multigroup invariance with AMOS), and gender invariance was evaluated using ΔCFI, which is the most sensitive index to detect a lack of invariance (Meade et al., 2008), employing the absolute value of ΔCFI equal or lower than .01 (Byrne, 2012;Cheung & Rensvold, 2002). We also considered a change ≤ .015 in RMSEA as an indicator of invariance (Chen, 2007). As for validity, we examined convergent validity by computing bivariate correlations between HSS scores and scores on the ASI (Glick & Fiske, 1996) and AMI (Glick & Fiske, 1999). Bivariate correlations were also computed between HSS scores, age, and religiosity. Religiosity was coded as a dummy variable, with "non-religious" as a reference.

Item Analysis
Means, standard deviations, skewness, and kurtosis for each HSS item in the first subsample (n = 1394, F% = 50.4, age M = 22.28, SD = 3.44) were calculated. They are shown in Table 1. All except three items (item 3, item 7, and item 22) showed skewness and kurtosis within normal parameters (being included in the conventional cutoff of ± 3 [e.g., Mayers, 2013]). Regardless of statistical significance, simulation studies have found that serious problems may exist when univariate skewness is ≥ 2.0 and kurtosis is ≥ 7.0 (Curran et al., 1996).

Exploratory Factor Analysis
Based on items' distribution, average correlation with other items, and item-total correlations (Clark & Watson, 1995), this data was suitable for factor analysis. The Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy (.91) and the significance of Bartlett's test of sphericity of the present study (χ 2 (231) = 8999.047, p < .001) indicated that this data had adequate factorability. The results of the factor analysis indicated that four factors -which explained 48.98% of the total variance -should be extracted. However, as shown in Table 2, some items (i.e., item 7, item 16, and item 22) showed poor factor loading (< .30) and item 4 showed saturation on more than one factor and the difference between their primary and alternative factor loadings was not within 0.20 (Howard, 2016). Therefore, we eliminated these items and performed again the exploratory factor analysis.
The results of the factor analysis indicated that four factors -which explained 53.38% of the total variance -should be extracted. As shown in Table 3, all 18 items in the HSS had good factor loadings. Cronbach's alphas had adequate values (see Table 3). The four factors identified were partially different from those of the original version. In particular, we identified a first factor that we named "Courtship Strategies", which included all the items of Factor 3 of the original HSS (except item 16, which was removed), three items (item 1, item 2, and item 8) from Factor 1 of the original HSS, and one item (item 12) from Factor 3 of the original HSS. The second factor that was identified was equal to Factor 2 of the original HSS, except that it included item 3 (which originally belonged to Factor 1), and it did not include item 12; we named this factor "Women Depend on Men". The third factor was composed of two items (item 5 and item 6) which, in the original HSS, belonged to Factor 1; we named this factor "Women's Self-Sacrifice". The fourth factor was composed of the same

Confirmatory Factor Analysis
To verify the factor structure identified through EFA, a CFA was performed on the second subsample (n = 1387, females = 50.3% M age = 22.35, SD = 3.61). The results showed that the fit indices were not completely acceptable (CFI = .891, TLI = .871, RMSEA [90% CI] = .064 [.060; .068], SRMR = .051). Modification Indices (MIs) suggested adding error covariance between items 1 and 2, and between items 13 and 14, both couples of items belonging to the Courtship Strategies dimension. Scrutiny of the content of each of these items revealed a substantial overlap between item 1 "The best way for a girl to attract a boyfriend is to use her body and looks" and item 2 "There is nothing wrong with men being primarily interested in a woman's body", and among item 13 "Being with an attractive partner gives a guy prestige" and item 14 "Guys who are able to date a lot of people (players) are considered cool". This overlap in item content can lead to error covariances (Byrne, 2004 Standardized factor loadings ranged from .42 to .79 and were significant at the .001 level. The correlations between the four factors were all significant at p < .001, positive, and comprised between .42 and .77 (Fig. 1).

Gender Invariance
Gender invariance analyses were conducted with 685 males and 702 females. As a prerequisite, we tested the final four-factor model separately in males and females (Byrne, 2004), using AMOS (Arbuckle, 2007   In line with the recommended practice for testing measurement invariance (Dimitrov, 2010;Little, 1997;Vandenberg & Lance, 2000), first, the independence model was fitted (χ 2 = 6003.22, df = 72, p < .001). As reported in Table 4, configural invariance and weak or metric factorial invariance were supported, confirming that the factor loadings were equal across genders. Then, scalar or strict invariance, which constrained intercepts to be invariant across groups, was tested. However, neither scalar invariance nor partial scalar invariance was demonstrated.

Convergent Validity
We examined construct validity by considering bivariate correlations between HSS scores and scores on the AMI and the ASI. As reported in Table 5, all HSS factors were significantly and positively associated with benevolent and hostile sexism toward both men and women, which indicates that individuals who more strongly endorse the norms of the Heterosexual Script hold more sexist beliefs. No association was found between overall endorsement of the Heterosexual Script and age; however, a negative correlation was found between age and the "Women's Self-Sacrifice" factor of the HSS (correlation was significant at level .05). Being religious was positively and significantly associated only with the "Women Depend on Men" and the "Sex Defines Masculinity; Women Set Sexual Limits" subscales, and with overall Heterosexual Script endorsement.

Discussion
The Heterosexual Script Scale has not yet received scientific attention in Italy, one of the countries in which gender stereotypes are particularly prominent. The purpose of this study was to examine adherence to sexual scripts among Italian young adults by first evaluating the psychometric properties of the Italian version of the Heterosexual Script Scale (HSS, Seabrook et al., 2016). The factor structure, internal consistency, convergent validity, reliability, and gender invariance of the Italian HSS were tested, using data collected from a sample of young Italian women and men. The original HSS (Seabrook et al., 2016) was developed to investigate gender scripts and stereotypes concerning commitment, courtship, and sexuality in the context of heterosexual relationships in young adults. The original HSS counts 22 items, which are articulated into four subscales: "Courtship and Commitment" (Factor 1), "Men as Powerful Initiators" (Factor 2), "Men Value Women's Appearance" (Factor 3), and "Sex Defines Masculinity, Women Set Sexual Limits" (Factor 4).
An exploratory factor analysis was performed to investigate whether the Italian HSS presents the same factorial structure as the original scale. Results indicated that three items from the original scale (item 7, item 16, and item 22) showed poor factor loadings, and one item (item 4) showed saturation on more than one factor. Therefore, these items were removed, and a second exploratory factor analysis was performed. Results indicated that four factors should be extracted. Next, a confirmatory factor analysis was performed, which indicated that the 18-item four-factor structure of the Italian HSS showed a good fit to the data.
The current findings indicate that the four factors of the Italian HSS do not correspond to the original HSS subscales. Indeed, in our sample, Factor 1 and Factor 3 from the original scale merged, thus creating a unique dimension that investigates courtship strategies for men and women and beliefs about the importance of women's physical appearance. We named this factor "Courtship Strategies". The fact that these separate sets of beliefs were found to belong to the same dimension seems to suggest that, for the population of young Italian adults, those two aspects are closely related. As the importance of physical attractiveness is particularly emphasized in Italy (Di Giacomo et al., 2018;Policardo et al., 2018), young Italian people may consider physical appearance as a key aspect that contributes to courtship and romantic commitment. More specifically, the belief that men value women's physical attractiveness may, by itself, be considered an important determinant for romantic relationships; however, further investigation would be needed to confirm this finding. The second factor that was found for the Italian HSS comprises three items from the "Men as Powerful Initiators" subscale of the original HSS, plus one item from the "Courtship and Commitment" subscale (i.e., item 3, "No matter what she says, a girl isn't really happy unless she is in a relationship"). This factor globally assesses the belief that women need to be in a relationship with a man to feel fulfilled and protected (i.e., women are assigned to a passive role) while men are responsible for their partner's safety and for the sexual life of the couple (i.e., men should behave as active initiators); we named this factor "Women Depend on Men". The fact that item 3 was found to belong to the "Women Depend on Men" subscale of the Italian HSS suggests that, in the Italian context, the belief that women need to be protected by men is closely related to the belief that women cannot be happy when they do not have a partner.
Next, a new factor that does not exist in the original HSS was extracted. This third factor comprises only two items (item 5 "Sometimes girls have to do things they don't want to do to keep their boyfriend happy" and item 6 "A woman should be willing to make personal sacrifices in order to satisfy her partner"), which specifically assess the belief that women should be willing to make personal sacrifices to please their partners. We named this factor "Women's Self-Sacrifice". The fact that item 5 and item 6 stand alone as a specific factor in the Italian HSS seems to indicate that, in the Italian context, the belief that women need to make personal sacrifices to satisfy their partner is a rather independent and well-defined concept, which does not belong to a bigger cluster of gender stereotypes. More precisely, the result suggests that women's self-sacrifice is not considered as one of the numerous strategies that women can implement to prove their commitment to the relationship, but rather as a specific and established feature of women's role that exists per se. Considering the undeniable influence of the Catholic Church on Italian culture, and given that the Catholic Church has a rather patriarchal structure (Accati, 1995), it is plausible to assume that the value that is attributed to women's sacrifice in Italy is, at least in part, a reflection of the model of women that is being promoted by the Church (i.e., the compliant Madonna). Moreover, as noted by Hunt et al. (2015), women's rate of employment in Italy is lower than that in other Western Countries and the gender pay gap is larger, which probably contributes to making women feel more dependent on men and more prone to make personal sacrifices to maintain the relationship. Indeed, when Hunt et al. (2015) compared the level of endorsement of different gender norms among Italian and Australian women, they found that Italian women tended to attribute significantly greater importance to the maintenance of romantic relationships compared to Australian women.
Finally, the Italian HSS comprises a fourth factor that includes all the items from the fourth HSS subscale, except for item 22, which was removed. We named this factor "Sex Defines Masculinity; Women Set Sexual Limits" (its original name), as it evaluates the belief that men are, by nature, always willing to engage in sex, while women instead are more interested in the emotional aspects of a relationship, and they should prevent sex from happening to fast. As mentioned above, item 4 "Girls should do whatever they need to (e.g., use makeup, buy attractive clothes, and work out) to look good enough to attract a date/partner" was removed because it showed saturation on more than one factor. We believe this was because the importance of physical appearance (including clothing, makeup, and body shape) is particularly emphasized in Italy and therefore pervades different aspects of people's romantic life (Policardo et al., 2018). Moreover, we removed item 7 ("Guys like to play the field and shouldn't be expected to stay with one partner for too long"), item 16 ("It's only natural for a guy to make advances on someone he finds attractive"), and item 22 ("Women with a lot of 'experience' should expect a bad reputation") because they showed poor factor loadings. Item 7 and item 22 showed low factor loadings also in the original validation study (Seabrook et al., 2016). Moreover, it is noteworthy that item 7, item 16, and item 22 address a common concept: the fact that it is natural for men to act on their sexual desires (even regardless of their romantic status), while the same is not appropriate for women. This belief may not resonate with young Italian adults, either because they judge that these behaviors are not acceptable neither for men nor for women or because they judge that women deserve the same sexual freedom that is accorded to men. The latest hypothesis would be consistent with the findings of Minello et al. (2020), who observed that the sexual attitudes of young Italian adults are evolving toward a seemingly more egalitarian direction. Alternatively, it is possible that the Italian version of these items was not sufficiently clear and therefore misled participants' answers. Further investigation would be needed to explore the reasons underlying this result.
Likely, the discrepancies that were found between the factor structure of the Italian HSS and the original HSS were 1 3 due to cultural differences between the samples. Indeed, as noted by Hunt et al. (2015), even though different countries share some common beliefs about women's roles, the assumption that gender scripts are uniform among Western countries is erroneous and should be avoided.
Despite differences in the factor structure, the Italian version of the HSS presents adequate psychometric properties in terms of both reliability and convergent validity. The internal consistency of the four subscales was found to be adequate to good. The third subscale had a lower internal consistency than the other factors, but this is not surprising as Cronbach's alpha tends to decrease as the number of items on the scale diminishes.
The construct validity of the Italian HSS was assessed by observing correlations between HSS scores, age, religiosity, and scores on two measures of sexism against men and women. As expected, all HSS factors were found to be significantly and positively correlated with both benevolent and hostile sexism toward men and women. The result is consistent with the findings of Seabrook et al. (2016) and confirms the theory suggesting that higher levels of endorsement of traditional gender scripts are associated with higher levels of sexism.
HSS total scores were not found to vary as a function of age, which suggests that the level of endorsement of gender scripts tends to remain relatively stable across different age groups. However, in our sample, younger women were more likely to endorse the belief that women should make personal sacrifices to maintain a romantic relationship, compared to their older counterparts. This finding might suggest that through time and experiences women progressively distance themselves from stereotypical expectations and learn to prioritize their own needs. However, as cultural gender scripts are mutable and dynamic, different generational cohorts may have been exposed to different cultural scripts; therefore, more research featuring different age groups would be needed to investigate the relationship between age and gender scripts' adherence.
In our sample, being religious was found to be associated with a stronger overall endorsement of the heterosexual script. In particular, religiosity was associated with a greater endorsement of the "Women Depend on Men" and the "Sex Defines Masculinity; Women Set Sexual Limits" subscales. These results are not surprising considering that numerous previous studies established an association between religious attitudes and traditional gender roles (Morgan, 1987); sexism (Glick & Fiske, 1997;Maltby et al., 2010;Mikołajczak & Pietrzak, 2014); and endorsement of the sexual double standard (Emmerink et al., 2015;Migheli & Pronzato, 2020). The fact that the "Courtship Strategies" subscale was not significantly associated with religiosity might be due to the fact that this subscale contains many items that assess women's sexualization and the belief that "it is normal for men to 'play the field'". These beliefs are probably not in line with the Catholic moral, since Catholicism tends to promote sexual conservativism.
The current study also aimed to assess the invariance of the HSS across genders. Assessing the gender invariance of any psychological measurement is important, as it provides evidence that the construct in question is being measured equally in males and females. This allows, for example, to reliably compare the scores of male and female respondents. Given that the heterosexual script is composed of different sets of norms for men and women, it was particularly important to assess whether men and women interpret and respond equally to the HSS. In our sample, gender invariance analyses provided support for configural invariance and metric factorial invariance, i.e., the construct is associated with the same set of items in each group (i.e., configural invariance), and the relationship between the construct and the items, as represented by factor loadings, is not significantly different across group variables (i.e., metric invariance). However, scalar invariance was not supported, indicating that the factor pattern coefficients and the intercepts are not equal across the gender groups. Thus, making mean comparisons across gender groups is not adequate as the absence of scalar invariance does not guarantee that mean differences in scores across the groups are real differences in the measured construct (Vandenberg & Lance, 2000). However, the absence of full metric invariance across genders is overall consistent with Seabrook et al. (2016) who found that Factor 1 and Factor 3 of the HSS had slightly different meanings for men and women (i.e., the full metric invariance of the HSS was not supported).
The present study features some important limitations. Firstly, most participants were recruited through the means of an Instagram profile that publishes content related to sexual education. This might represent a selection bias as people who follow a sex education page on Instagram might hold certain characteristics (e.g., open-mindedness) that do not fully reflect the general target population. Secondly, our study sample was composed of exclusively young adults; thus, our findings cannot be generalized to adolescents or older adults. Thirdly, as the HSS focuses on the endorsement of gender scripts in the specific context of heterosexual relationships, the present findings cannot be generalized to individuals who are not involved in romantic relationships or to individuals with different sexual orientations. Fourthly, the construct validity of the Italian version of the HSS was only tested by assessing Pearson's correlations with measures of benevolent and malevolent sexism. Assessing the associations with other theoretically related constructs, such as gender role endorsement and self-sexualization in women, would have provided further empirical support toward the construct validity of the Italian version of the HSS; however, because of difficulties in identifying and obtaining validated measures of such constructs in the Italian language, this was not possible in the current study. Future studies should investigate the associations of the Italian version of the HSS with other convergent and/or divergent validity measures to further test the construct validity of the scale. Finally, further studies featuring individuals in different age groups and cultural backgrounds are also warranted to support the validity of the present findings.

Conclusion
Despite the aforementioned limitations, the findings of the present study indicate that the HSS can be used to reliably assess the level of endorsement of gender scripts in the context of heterosexual relationships in young Italian men and women. As sexual double standards and sexism are relevant problems in the Italian context (Migheli & Pronzato, 2020;Rollero & Tartaglia, 2018), more research is warranted to investigate the effects that gender stereotypes may have on multiple dimensions of young people's well-being, including relational and sexual wellbeing.

Social Policy Implications
Overall, the findings of the current study indicate that the norms of the heterosexual script still influence the sexual attitudes of young Italian adults. Recent evidence shows that endorsing the norms of the heterosexual script is associated with numerous indicators of reduced sexual health (e.g., reduced sexual functioning and sexual assertiveness, Scappini & Fioravanti, 2022). To prevent these negative outcomes and cut their treatment costs, relevant authorities should consider implementing more educational interventions in schools. For instance, sexuality education curricula could be improved by integrating modules covering egalitarian sexual scripts. Furthermore, more research is warranted to gauge the efficacy and impact of such preventive interventions. In particular, longitudinal studies would be needed to assess whether and how education curricula could prove useful in reducing the level of endorsement of the heterosexual script, from adolescence to young adulthood.
For what concerns college students, previous research has shown that exposure to the heterosexual script and endorsement of sexist beliefs are associated with higher rape myth acceptance (Angelone et al., 2020;Hust et al., 2016), especially in young men (Martini et al., 2021). Considering the high level of endorsement of the heterosexual script that was found in the current sample, it appears highly relevant to develop and implement state-level prevention programs to target sexist and gendered beliefs on college campuses. Likewise, rape-prevention programs could be offered in other settings involving young adults, such as private and public work environments, to reach more segments of the population. Furthermore, considering that social media are major channels of communication, especially for young adults, we believe that more efforts should be put into trying to contain the spread of gendered and sexist messages on popular social media platforms. Likewise, more attention should be put to close the large gender equality gap that still exists in traditional media (Global Media Monitoring Project, 2020), as this could help promote more equalitarian gender views.
Finally, the current findings seem to suggest that, although the heterosexual script exists in all Western Countries, some culture-specific variations might exist, with some beliefs being more prevalent in certain cultural contexts. For this reason, researchers and sexual health practitioners should be mindful of context-specific sexual scripts when designing research and intervention protocols.

Appendix. Items of the Heterosexual Script Scale (HSS) (Italian version in brackets)
1. The best way for a girl to attract a boyfriend is to use her body and looks. (Il miglior modo per una ragazza per attirare un fidanzato è usare il proprio corpo e il proprio aspetto fisico.) 2. There is nothing wrong with men being primarily interested in a woman's body. (Non c'è niente di male nel fatto che gli uomini siano principalmente interessati al corpo di una donna.) 3. No matter what she says, a girl isn't really happy unless she is in a relationship. (A prescindere da quello che dice, una ragazza non é davvero felice finché non ha una relazione.) 4. Girls should do whatever they need to (e.g., use makeup, buy attractive clothes, and work out) to look good enough to attract a date/partner. (Le ragazze dovrebbero fare tutto quello di cui hanno bisogno per essere abbastanza belle (per esempio truccarsi, comprare vestiti attraenti ed allenarsi) per attirare un partner.) 5. Sometimes girls have to do things they don't want to do to keep their boyfriend happy. (Ogni tanto le ragazze devono fare cose che non vorrebbero fare per rendere felici i loro ragazzi/fidanzati.) 6. A woman should be willing to make personal sacrifices in order to satisfy her partner. (Una donna dovrebbe essere disposta a fare sacrifici personali per soddisfare il proprio partner.) 7. Guys like to play the field and shouldn't be expected to stay with one partner for too long.
(Ai ragazzi piace darsi da fare e non ci si dovrebbe aspettare che rimangano con la stessa partner per troppo tempo.) 1 3 8. Women are attracted most to a man with a lot of money. (Le donne provano più attrazione per gli uomini che hanno molto denaro.) 9. A man should always protect and defend his woman. (Un uomo dovrebbe sempre difendere e proteggere la sua donna.) 10. Men should be the ones to ask women out and initiate physical contact. (Dovrebbero essere gli uomini a chiedere alle donne di uscire insieme e ad iniziare il contatto fisico.) 11. A woman wants a man because she wants someone to protect her. (Una donna vuole un uomo perché vuole qualcuno che la protegga.) 12. Women like to admire men's bodies and are attracted most to men who are muscular and handsome. (Alle donne piace ammirare il fisico degli uomini e sono attratte soprattutto da uomini muscolosi e affascinanti.) 13. Being with an attractive partner gives a guy prestige. (Avere una partner attraente conferisce prestigio ad un ragazzo.) 14. Guys who are able to date a lot of people (players) are considered cool. (I ragazzi che riescono ad avere tante ragazze sono considerati fighi.) 15. In the dating game, guys frequently compete with each other for partners, and girls try to lure or catch partners. (Durante il corteggiamento spesso gli uomini competono gli uni con gli altri per una ragazza mentre le ragazze cercano di attirare e adescare gli uomini.) 16. It's only natural for a guy to make advances on someone he finds attractive.
(Solo per gli uomini è naturale fare avance a qualcuno che trovano attraente.) 17. It is natural for a guy to want to admire or check out other people, even if he is dating someone. (È normale che un ragazzo voglia guardare o ammirare altre ragazze anche quando è fidanzato.) 18. Guys are always ready for sex. (Gli uomini sono sempre pronti per il sesso.) 19. Most guys don't want to be 'just friends' with a girl. (La maggior parte dei ragazzi non vogliono essere 'solo amici' con una ragazza.) 20. Guys are more interested in physical relationships and girls are more interested in emotional relationships. (I ragazzi sono più' interessati alle relazioni fisiche, le ragazze sono più' interessate alle relazioni emotive.)

21.
It is up to women to keep things from moving too fast sexually. (Sta alle donne far sì che le cose non succedano troppo velocemente dal punto di vista sessuale.) 22. Women with a lot of 'experience' should expect a bad reputation.