“Tell Me who You Hang Out with and I'll Tell You Who You Are”: Analysis of How the Social Contacts´ Gender Predicts Self-Concept as a Socially Engaged Person and Intention of Future Social Participation Among Young People

The purpose of this work was to study the effect of gender on the social influence processes that explain the young people’ social participation. A total of 962 Spanish undergraduate students completed a questionnaire assessing their likelihood of engaging in social participation in the future, their self-concept as socially engaged people, and the number of socially involved contacts in their social networks, distinguishing between male and female contacts. We found that gender homophily characterised the social networks of both women and men. Consistent with gender stereotyping, being female predicted self-concept and the intention to be civically engaged

nterest in studying and identifying the gender gap in many different areas and from different perspectives has increased progressively since the 1990s.Studies that have addressed gender differences have done so in relation to many different topics and across a wide variety of behavioural and attitudinal dimensions.In turn, these findings have sparked considerable interest in identifying the processes responsible for the differences found and the conditions under which they occur (Vogel et al., 2003).Knowing these processes and conditions could provide important clues to understanding the positions that women and men hold in society and developing intervention policies and strategies that promote social equality and access to the same opportunities.
Social participation is one of the areas in which gender differences have been examined.Social participation is an extremely broad concept that alludes to a wide range of actions, and this has led to it being defined in several different ways.In this research, we consider it as the behavioural component of active citizenship (Enchikova et al., 2019), which includes actions of participation in social, political or community life, characterised by mutual respect and nonviolence, in accordance with human rights and democracy (Hoskins & Mascherini, 2009).
Studying young people's social engagement and the gender dimension of it is crucial because youth is a critical period for developing patterns of participation that are maintained throughout life (e.g., Greenfield & Moorman, 2018).On the one hand, at present, there seems to be a consensus in the literature that the young are less involved in representative politics, such as joining a political party (e.g., Kiesa et al., 2007) and that they prefer another type of involvement, based on individual commitment and less structured forms, such as activism or volunteering (e.g., Bennett, 2007).However, in any case, social participation has numerous benefits at all levels, including well-being and empowerment (e.g., Cicognani et al., 2015).On the other hand, while much of the research on the gender gap has focused on adults, less attention has been paid to youth or emerging adults (Stefani et al., 2021).For this reason, it is important to identify the factors that can promote social participation at this vital stage and work to reduce any gender gap that may increase later in adulthood.
Previous research on the factors that explain the gender gap in participation has focused primarily on political knowledge or political efficacy among women and men (e.g., Wen et al., 2013).At the same time, despite the growing interest in studying the impact of social influence, and specifically of social networks on social participation (e.g., Tseng & Hsieh, 2015), gender perspective has so far scarcely been incorporated in this type of research.Thus, the characteristics of networks of women and men may be different, but it is largely unknown whether and how this circumstance play a role in explaining the differences in their social participation (Dávila et al., 2020).
With these considerations in mind, the main objective of this work is to study the effect of gender on the social influence processes that explain social participation in young people.

Gender Differences in Social Participation
Women's social participation has increased in recent decades thanks to their better education, higher social status and the use of new information and communication technologies, among other factors.However, significant differences between women and men remain (Gottlieb, Wen et al., 2013) and are also observed among young people (e.g., Dávila et al., 2020).Some authors suggest that although the differences may seem rather small during adolescence and young adulthood, they become more pronounced in adult life (Malin et al., 2015).As previously described, the younger generation's interest in politics is generally lower than that of the previous generation and young people are more attracted to voluntary associations, social movements, and other civic groups (Albanesi et al., 2012).However, within this general pattern, it is evidenced that young women are more prone toward civic or community action, while young men are more involved in political participation (Stefani et al., 2021).Civic or community participation refers to activity that focuses on helping others and supporting the common good (e.g., volunteering) and political participation comprises actions aimed to influence political outcomes or bring about social change (e.g., membership in a political party) (Barrett & Brunton-Smith, 2014;Stefani et al., 2021).Each type of participation has different purposes, is promoted by different factors (Pavlova & Silvereisen, 2015), and may have different implications for achieving gender equality.
Some theoretical explanations for gender differences in social participation have focused on the role of unequal resources and opportunities (access to education, employment, etc.), while others have emphasised social processes such as gender socialisation (Stefani et al., 2021).In this regard, Social Role Theory (Eagly, 1987) may help explain the influence of gender roles or stereotypes on the social participation of women and men.Gender roles specify what kind of behaviour is expected from an individual based on his or her sexual identity.Thus, males are encouraged to be more agentic (assertive, controlling and independent, for example) and females are expected to be more communal (interested in the well-being of others, more sensitive to interpersonal relationships, etc.).Therefore, women and men may behave differently because the social roles they play are associated with different expectations and require different skills (Cicognani et al., 2012).
These theoretical foundations also allow us to understand the results of several studies indicating that gender differences in social participation derive from other differences in political knowledge, political efficacy, influence of parents' political participation, etc. (see for example, Cicognani et al., 2012;Gordon, 2008;Wen et al., 2013).Specifically, the work of Cicognani et al. (2012) shows the differential role of social influence.They demonstrate that the political participation of adolescents is influenced by parental participation, especially in the case of females.This would be consistent with the content of the stereotype that defines women as more sensitive to interpersonal relationships.Along the same lines, Dávila et al. (2020) have shown that social ties play an important role in explaining the intention for social participation and that this varies with gender.In this sense, they demonstrate that the composition of networks is significant in predicting women's civic participation and men's political participation.However, while those studies address the gender dimension by analysing the type of relationship (e.g.strong and weak ties), the present work focuses on the characteristics of social contacts, specifically their gender, as described below.

Gender Dimension in Social Networks
It is noteworthy that, although previous research has evidenced the role of social networks in explaining social participation (e.g.Tindall, 2002), there has been little incorporation of a gender perspective in their analysis.However, its introduction may shed light on the dynamics on how people get involved.For example, having a larger network increases the chance of being asked for donations or to volunteer (Passy, 2003).In this sense, the size of the social networks that characterise people of one gender may lead to an increase or decrease in the likelihood of them contributing socially.Additionally, women and men may differ in the probability of accepting invitations to engage (De Wit & Bekkers, 2015).For example, given female gender role stereotyping, women may be more likely to accept requests for collaboration, as they are more responsive to requests for helping behaviour (Croson & Gneezy, 2009).Although gender differences in relation to network size are not clear-cut (see for example, Smith, 2000), they have been found in other aspects of the structure and composition of networks (from social network analysis referred to as alter attributes and ties between alters; see in McCarty et al., 2019).For example, women and men differ in the number of strong and weak ties present in their social networks (Smith, 2000).While the impact of tie strength on social participation has been shown in several studies (Dávila et al., 2020;Tindall, 2002;Son & Lin, 2008, for example), other aspects of social network composition, such as the gender of the contacts, have received little attention.
Previous research has shown that social networks of women and men have been found to be characterised by homophily with respect to gender (Almquist et al. 2014;Friebel et al., 2021;Powazny & Kauffeld, 2021).Homophily refers to the tendency of people to connect or interact with peers who are like them, which could be explained by the similarity-attraction hypothesis (Powazny & Kauffeld, 2021).But what implications might gender homophily in social networks have when explaining social participation?On the one hand, Smith et al. (1999) argue that certain contextual factors, such as a greater presence of other women or men, may have an impact on this.That is, it may make one gender more salient in these social networks and thereby activate the gender role content ascribed in each case, encouraging more gender-role behaviours.Others may reinforce gender-typical behaviour when gender is salient within a context.The results of the study by Athenstaedt et al. (2004) are also along these lines.On the other hand, it has been suggested that demographic homophily could be one of the main drivers of trust relationships and increase social bonding in the social network.It would also act as a trigger for information diffusion processes and, therefore, for social influence processes (Ahlf et al., 2019).
Added to this, the effects of social networks need not be the same for women and men.For example, Lilleker et al. (2021) found that women with larger virtual networks were more likely to engage with political information in comparison to men who had a similar-sized network on a particular platform.Previous empirical evidence has shown that women tend to be more attuned to social cues than men (Djupe et al., 2007).Women place more value on relationships with others, so it is to be expected that friends and acquaintances play a more important role in women's social participation (Wen et al., 2013).Wen et al. found that in the specific case of political participation, women responded to both internal and external factors, while men were more motivated by internal factors such as confidence in their political capability.This moderating role of gender has also been found in relation to the impact of other variables on social participation, such as civic knowledge or efficacy beliefs, for example (Manganelli et al., 2014).
Finally, the characteristics of the social networks could shape how we see ourselves and, in turn, promote social participation.Current theoretical approaches view identity as a social achievement in progress, and as such, as both a pathway to and a result of social relations (O'Toole et al., 2016).Social networks convey meanings (through rituals and narratives, for example) that build and solidify identities (Passy, 2003;Somers, 1992).Dávila et al. (2021) showed that acting consistently with seeing oneself as a committed person functioned as a significant mediator of the influence of social networks in predicting volunteering, which would be in line with the role-identity model of volunteering proposed by Piliavin and colleagues (e.g., Charng et al., 1988).Once a certain role is assumed as part of the personal identity, this promotes the development and implementation of that role, as the person finds reinforcement in performing behaviors consistent with his or her identity.
In this case, gender may also have something to do with the influence of social ties in shaping our self-concept.A considerable amount of research has focused on whether women and men differ in their self-concept.It has been found that women are encouraged to develop a self-concept in which relationships with others are most important.In contrast, men are encouraged to develop a self-concept in which independence from others is most important (Jackson et al., 1994).
Based on the above, the first specific objective of this research is to identify the characteristics of the composition of the social networks of women and men in terms of the presence of socially engaged individuals in them.The second objective is to determine to what extent gender and the composition of the social networks play a significant role in predicting self-concept as socially engaged individuals and also in predicting future social participation.Our third goal is to study whether the connections between social network composition, selfconcept and intention for future participation are moderated by gender.In this way, we aim not only to determine the mechanisms that explain social participation, but also to identify the extent to which these mechanisms may be moderated by gender.Figure 1 shows the relationships that will be studied in relation to the latter objective.

Figure 1 Analysis Model
Given the possible differences in relation to the prediction of civic and political future participation, the proposal is to conduct the analysis for each type of participation independently.
The hypotheses to be tested are described below: Previous studies have found that women and men have more same-sex contacts in their networks than opposite-sex contacts (Friebel et al., 2021), and this leads us to expect that we would find the same pattern with respect to socially engaged contacts in the study participants' networks.
• Hypothesis 1.The gender composition of social networks will be different for women and men.➢ As previously described, the composition of social networks shapes how we see ourselves (Passy, 2003).Similarly, adopting a certain gender can lead to us seeing ourselves in the light of the gender role stereotypes as a result of the socialisation process (Eagly, 1987).In line with it, different studies have shown that women and men are more attracted to different forms of social participation (Stefani et al., 2021, for example).• Hypothesis 2. The social network composition and gender of study participants will be significant predictors of both their self-concept and the type of social participation.
➢ Several studies have shown the moderating role that gender has on the impact of certain variables, such as efficacy beliefs, for example, in predicting social participation (Manganelli et al., 2014).In addition, other empirical evidence has demonstrated a greater impact of social contacts on women than men (Cicognani et al., 2012).• Hypothesis 3. The gender of the participants will moderate the relationship between both social network composition and social participation intention, as well as the relationship between social network composition and self-concept.

Participants and Procedure
The participants were 969 social science university students at a public university in Madrid, Spain.One person was missing data and six participants indicated their non-binary status.
Given the small number of non-binary persons, it was decided not to consider them for the analyses carried out.Finally, of the remaining 962 participants, 73.18% were women and 26.81% men.Ages ranged from 18 to 30 years, with a mean of 20.41 years (SD=2.28).
The mean age of the men was slightly but significantly higher (T=3.43;p<.01) than that of the women (M=20.84,M=20.25, respectively).No significant differences by gender were found in previous experience in social participation (Pearson's chi-squared=.81;p>.05), or in the total number of contacts they had with social participation experiences (Z=-.37;p>.05).
Data were collected between November 2018 and January 2019.Participants were contacted during class time and were selected using a non-probabilistic procedure.The participation was voluntary.All participants signed an informed consent document where they were informed of the study's aims and assured that their data would be treated with total confidentiality.They then completed a pen-and-paper questionnaire.

Instruments
The questionnaire included the following measures: Likelihood of Future Participation.In this research, we propose to study the intention to participate socially as a proxy for actual participation.As indicated by the theory of reasoned action (Ajzen and Fischbein, 1980) and the theory of planned action (Ajzen and Madden, 1986), behaviour is directly determined by the intention to execute it.A number of different studies show the connection between both elements and the predictive power of behavioural intention (e.g., Lee et al., 1999).
Together with the items of the scale previously used by Dávila et al. (2020;2021), which allows estimating the probability of becoming involved in five types of social participation in the near future (membership in an association with social purposes; volunteering in a non-profit organization; membership in a political party; membership in a popular unity candidacy; and activism in a social movement), the probability of becoming involved in "free participation without establishing commitments with social entities" and "social entrepreneurship" was evaluated, considering that these two additional behaviors could describe emerging forms of participation in young people.The responses were given on a 7-point Likert-type scale ranging from 1 'not likely' to 7 'very likely'.To summarize the information provided by the items described and simplify the analysis of the relationships, an exploratory factor analysis (principal components, varimax rotation) was performed, which showed that the items were organized into two factors that explained 61.07% of the variance.The first factor was labelled "political participation" (membership of a political party and membership of a popular unity candidacy).The second factor was labelled "civic participation".In both cases, acceptable reliability values were found (.78 and .71, respectively).
Social Networks.We used a name generator and focused on a specific subset of the participants' network that was relevance to our research question.Specifically, we asked participants to provide a list of the people they knew who were engaged in any type of civic and political participation.The gender of each of the identified alters (social contacts) was deduced from their names.If the name could allude to either the male or the female gender, or it was considered impossible to determine the associated gender, it was coded as a missing data (it was not possible to identify the gender of 18 social contacts).We calculated the following simple compositional variables from personal networks: • Number of male contacts • Number of female contacts • Index of qualitative variation (IQV) to measure gender heterogeneity.This index is a measure of variability for nominal variables, such as gender.It is based on the ratio of the total number of differences in the distribution to the maximum number of possible differences in the same distribution, therefore, it is an index based on the ratio of females to males or vice versa (for the calculation, see McCarty et al., 2019).The value 0 indicates complete homogeneity (contacts are all of the same gender) and value 1 indicates complete heterogeneity (contacts are spread evenly over the different genders, female and male).174 participants did not identify any alters or contacts in their social networks who were engaged in any kind of social participation, so the IQV to measure gender heterogeneity was not estimated in those cases.
Self-concept.Four items rated on a 7-point Likert-type scale (1 "strongly disagree" to 7 "strongly agree") were used to measure to what extent the participants saw themselves as socially engaged people (e.g."I see myself as a socially committed person"; "Being a socially engaged person is an important part of who I am, of my identity").Cronbach alpha was .91.Exploratory factor analysis (principal components) showed the existence of a single factor explaining 80.56% of the variance.

Data Analysis
Given that many of the variables being studied did not have a normal distribution (personal network measures often exhibit a non-normal distribution, see McCarty et al., 2019), we opted to use non-parametric techniques.Thus, the Mann-Whitney U test was used to compare the gender composition of social networks for women and men.We also conducted multiple regression analyses using the bootstrapping method (Efron & Tibshirani, 1993) to determine the impact on self-concept and on the two types of social participation of the variables linked to the composition of social networks.We performed moderated mediation analyses (conditional processes analysis) using Model 8 with a moderating variable (gender) in the PROCESS macro (Hayes, 2013) to determine the direct and indirect impact through the selfconcept of the variables related to the composition of social networks on the prediction of the intention to engage in social participation.Model 8 allows for the moderation of the direct effect as well as the indirect effect.

Results
The gender composition of the social networks linked to the social participation of women and men is shown in Table 1.It can be seen that there is a large dispersion in all the variables that reflect the composition of the social networks in both men and women.Women's social networks were characterised by a higher number of female contacts and a lower number of male contacts compared to those of the men.Similarly, men's social networks were characterised by a higher number of male contacts and a lower number of female contacts.Therefore, gender homophily was shown to characterise the social networks of women and men.Gender had a significant effect on both the prediction of self-concept and the prediction of both types of social participation (Table 2).However, that effect was different in each case.Thus, being female predicted seeing oneself as socially engaged and intending to participate civically, while being male predicted the intention to engage politically.The number of both female and male contacts predicted self-concept, with the number of female contacts being more important.However, only the number of male contacts predicted the intention to engage politically and only the number of female contacts predicted the intention to engage civically.The gender heterogeneity of contacts (IQV gender) only played a significant role in predicting civic participation, but its affect was weaker than the number of female contacts.Although the percentages of variance explained were small, Cohen's f2 (Cohen, 1988) is considered appropriate for calculating effect size so the effect sizes were medium.Table 3 shows the results derived from the moderated mediation analyses.For political and civic participation, both the direct and indirect effects of the social networks on the intention to engage were not moderated or determined by gender, with the exception of the indirect effect of the IQV gender.Only in this case was gender found to moderate the relationship between IQV gender and the perception of oneself as a socially engaged person.That is, for men, having a more gender-heterogeneous social network contributed to a greater extent to them seeing themselves as socially engaged than it did for women.With regard to the percentages of variance explained in each type of social participation, in general terms we could say that their effect sizes (Cohen's f 2 ; see Cohen, 1988) indicate a medium effect in the case of the intention to participate politically and a large effect in the case of civic participation.

Discussion
The issue of equal opportunities for women and men has been gaining greater attention in recent years and not just in the academic world.It is central for socio-political life in many societies and in some countries such as Spain its relevance is similar to that of the climate emergency (INJUVE, 2021).
In order to achieve these equal opportunities, we need to understand the different impacts that many factors have on women and men.Only in this way will it be possible to design proposals for interventions that really bring us closer to this gender equality.This research aims to address the analysis of a seemingly small element that can have very significant implications for other issues, generating a kind of domino effect.We set out to study how social influence processes may affect women and men in different ways, specifically by examining how the gender of socially engaged people in individuals' social networks influences their intention to participate socially.
Our first specific objective was to analyse the differences between women and men in terms of the gender composition of their social networks.The results show that such differences do exist and support the existence of gender homophily in this type of social contacts.This supported our Hypothesis 1.These results are in line with the findings in other fields of study (e.g., Friebel et al., 2021;Powazny & Kauffeld, 2021).
The second objective focused on analysing the role of social network composition and gender in predicting self-concept and social participation intention.The results support Hypotheses 2. On the one hand, our results support the idea that social contacts play a significant role in predicting the extent to which a person sees themselves as socially engaged.This coincides with what has been described by some authors who consider that self-concept is a continuously developing construct nurtured by the social relationships we establish, but at the same time guiding their establishment (Passy, 2003;O'Toole et al., 2016;Somers, 1992).However, our results also support the idea that it is important to consider the gender of the contacts in the analysis of their impact on self-concept.The number of females socially engaged contacts explains whether a person sees themselves as socially engaged to a greater extent.Given the cross-sectional design of our research, the reverse process could also be feasible, in which a person with a strong self-concept as socially engaged will seek out relationships with other socially engaged people to a greater extent, and as we have seen previously, being female is a predictor of seeing oneself in this way.If this were the case, this result would reflect another form of homophily based on how people see themselves and not so much on being one gender or the other.Future longitudinal research may disentangle this issue.
Also regarding Hipothesis 2, the finding that being female is a predictor of both self-concept as an engaged person and civic participation intention, and that being male is a predictor of political participation, fits perfectly with the content of gender role stereotypes (Eagly, 1987) and with some results from previous research also conducted with young people (Albanesi et al., 2012;Stefani et al., 2021).When considering the gender of social contacts, the number of male contacts was a significant predictor only of the intention for political participation, not civic participation.In contrast, the number of female contacts was a significant predictor of civic, but not political, participation intention.The greater presence of one gender or the other in social networks can be understood as a contextual factor that leads to the activation of one element of gender role content or another, and this in turn leads to the development of more behaviours, or in this case behavioural intentions, consistent with that content (Athenstaedt et al., 2004;Smith, 1999), probably anticipating greater social acceptance with it.
We have just seen that having one gender or the other is important in predicting intention for social participation, but even more important is the number of male or female contacts in a person's social network.This is the case both in predicting self-concept and in predicting the intention for social participation.When the gender variability of social contacts (IQV gender) is considered, it appears only as a significant predictor of civic participation intention, but it is less important than the number of female contacts and the fact of being a woman.Perhaps the practice and the social perception of civic participation are more flexible and diversified compared to those of political participation.That is to say, it is not perceived or even performed as an exclusive niche for female participation, whereas political participation is more reserved for men, also in terms of social perception.Perhaps for men it is more important to have this gender variability in their social contacts when it comes to establishing their intention to participate civically.Although this is not supported by the finding regarding the moderating role of gender on the direct relationship between the gender IQV effect and civic engagement intention, it would be consistent with the result regarding the finding of moderated mediation in the prediction of civic intention through self-concept.In predicting civic participation, gender heterogeneity in social networks would be more important for males to see themselves as socially engaged.In other words, our results indicate that having gender heterogeneity in the social network with respect to social participation more influential for men to see themselves as engaged people, and this produces a significant indirect contribution to them expressing a greater intention to volunteer, for example.
Lastly, with regard to the third objective proposed, concerning the analysis of the moderating role of gender in the relationships between the composition of social networks, self-concept and social participation intention, the results show that the mechanism linking these variables does not differ between women and men.In all cases, self-concept functions as a partial mediator of the relationship between social network composition and intention to participate socially, and the relationships are not moderated by gender.The only exception is the case discussed in the previous paragraph, where only the gender variability of social contacts was found to be more important for males than for females in predicting self-concept as a socially engaged person.This result perhaps makes perfect sense if we consider that seeing oneself as such, as a socially engaged person, does not fit the content of the male gender role.The presence of more women than expected, given the tendency towards gender homophily in men's social networks, would be beneficial in generating a self-concept that promotes social participation.But equally, it might also be expected that people who see themselves in this way would seek relationships that respond to another type of homophily more linked to how they are and see themselves as people, rather than to gender.In summary, Hypothesis 3 is not supported by the results found.
In any case, in general terms, it seems that having female contacts with experience of social participation in our social networks contributes positively to increasing the intention to participate civically, and this is more important for men than for women, to the extent that female contacts bring heterogeneity or variability to their social networks.Therefore, the practical strategy to promote civic participation could be summarised under the slogan "Connect with women".
Considerening the different types of social participation, the findings again highlight that already shown in previous studies (Dávila et al., 2020(Dávila et al., , 2021;;Pavlova & Silvereisen, 2015) in relation to the fact that the factors that predict them are not the same and that it is necessary to study the two types differently.Similarly, our finding that the size of the effect obtained by the moderated mediation model proposed in the explanation of civic participation is larger than in the case of political participation suggests the need to consider other explanatory mechanisms through the mediating role of other variables in the explanation of political participation.The study by Lilleker et al (2021) also shows the need to consider the online environments when addressing the gender gap.
In summary, taking into account the practical implications of the role of social influence through social contacts, it seems advisable in order to increase the mobilization of people to develop social participation to try to create spaces that promote that those who do not participate have the opportunity to network with people who already do.In more specific terms, women raising their profile in the world of politics and this, in turn, facilitating new changes in the direction of generating a more egalitarian society, requires an increase in their intention to participate politically.However, to achieve this, we need to break a "vicious circle": women's social networks are characterised by homophily in gender, and it is precisely the male contacts that promote political participation.Thus, any intervention that promotes an increase in the number of male contacts with experience in social participation in women's networks could increase their intention to participate in politics.These male contacts would contribute to generating a context in which participating in politics would be perceived as more desirable.At the same time, one should also not lose sight of the power that demographic homophily can have in creating stronger and more influential ties (Ahlf et al., 2019), perhaps promoting contact with women with a strong political commitment can contribute to this.In the past, interventions have often been focused on character education programmes, political simulations or service-learning initiatives (Lin, 2013).Our results contribute to new insights for steering interventions to foster civic engagement in other directions and through different means.
Regarding to the study's limitations, it is possible that relationships with other people may not only facilitate engagement, but may also block it.In our study, we only examined the role of social contacts who engage in social participation.However, it may be that the informants' relationships with individuals who have never engaged in social participation counteract the impact of those who have (McAdam & Paulsen, 1993).Furthermore, only the number of direct contacts in the social network has been considered, overlooking the connections between the person's contacts that are not directly linked to him or her.Consequently, this approach has underestimated the person's position in the network.Therefore, complete network structure has not been considered.Incorporating additional measures of the personal network, omitted from the current study design, could yield a more comprehensive and precise understanding (McCarty et al., 2019).
Another limitation is the use of intention rather than a measure of actual participation.While the prediction of a behaviour is mediated by the intention to perform that behaviour, following the theories of Reasoned Action and Planned Action, it would be desirable to develop future research that measures actual behaviour given the different factors that can lead to the intention behaviour gap (Sheeran & Webb, 2016).In addition, it would also be valuable to evaluate other forms of social participation, use other scales of measurement validated in the literature and to use other data collection techniques, such as in-depth interviews, that make it more feasible to study a person's entire network of contacts, their gender, the intensity of the relationships and the specific types of social participation they perform.In this way, it would be possible to have a more complete picture to analyse the interactions between all the elements of analysis that characterise the processes of social influence.Likewise, given the results obtained, it would be interesting to control for both adherence to gender stereotypes and the level of identification with one's own gender.Finally, it should not be forgotten that the research presented here is exploratory in nature, and is limited to a single correlational study using a student sample with a higher proportion of females than males.It would be necessary to carry out more studies with balanced samples concerning gender differences, with non-university students -there is potential for students to be more socially engaged than young people who are not in education -, even with an adult population, to determine to what extent the results can be generalized.
We are aware that our contribution is an isolated element in a complex web of interconnections between a large number of psycho-social, socio-economic and cultural factors that explain the experiences of women and men.However, despite this, we believe that it can help when proposing innovative interventions that contribute to making this world a better and more egalitarian place to live.In the same way, the approach and the results found also contribute to the development of a line of work that has been little considered to date, despite the growing interest in virtual social networks.The connection itself and what connects us with others plays an important role in how we decide to act.For example, in the case of some forms of social participation, their performance is presented as the result of individual choice (O'Toole & Grey, 2016), forgetting that actions and choices take place in dense social structures that constrain and enable them.

Table 1
Comparison of Social Network Composition by Gender

Table 2
Summary of Regression Analysis: Prediction of Self-Concept and Political and Civic Participation

Table 3
Moderated Mediation Summary