Influence of news media use and political discussions on social self-efficacy through sense of unity: an analysis of mediation model invariance with Spanish and Portuguese adolescents

Vol:.(1234567890) Current Psychology (2024) 43:9390–9403 https://doi.org/10.1007/s12144-023-04940-3 1 3 Influence ofnews media use andpolitical discussions onsocial self‑efficacy throughsense ofunity: ananalysis ofmediation model invariance withSpanish andPortuguese adolescents VanesaSalado1 · TaniaGaspar2 · ConcepciónMoreno‑Maldonado3 · MargaridaGaspardeMatos4 · FranciscoRivera1 Accepted: 30 June 2023 / Published online: 16 August 2023 © The Author(s) 2023 Abstract Discussing sociopolitical issues with family or friends during adolescence, as well as staying informed through media outlets, is key to developing active civic engagement for both youth and future adults. Likewise, these communicative exchanges promote a sense of belonging and social skills which foster adolescents’ wellbeing. This research aims to analyze the influence of participation in communication about political and social issues on adolescents’ social self-efficacy, both directly and as well as indirectly through sense of unity, in two Mediterranean countries (Portugal and Spain). The sample was selected through random multistage sampling by conglomerates. The 36,992 adolescents (50.6% girls and 49.4% boys) participated in the Health Behaviour in School-aged Children (HBSC) study in Portugal and Spain, and the age groups were distributed as following: 39.6% 13–14-years-old; 35.8% 15–16-years-old; and 25.6% 17–18-years-old. Analyses showed positive associations between the proposed indicator of civic engagement –news media use and political discussions– and social self-efficacy, both directly as well as through sense of unity, with similar results for adolescents in Portugal and Spain. Accordingly, educational programs promoting communication about sociopolitical issues could foster adolescents’ sense of unity and social skills, contributing to their positive development, wellbeing, and civic engagement. Curricular materials on current politics and social affairs could increase classroom dialogue based on respect, healthy peer relationships, and sense of belonging, fostering effective political socialization amongst youth and the development of democratic behaviors beneficial to society. Keywords Adolescents· Sociopolitical communication· News media use· Social self-efficacy· Sense of unity Introduction News media use and participation in political discussion with family and friends –along with sociopolitical interest, internal political efficacy, civic self-efficacy, etc.– are all considered to be indicators of adolescent civic engagement (Schulz etal., 2010). Despite discrepancies in the definition of the term, the majority of research underscores how civic engagement fosters positive youth development (Lerner etal., 2005; Sherrod, 2007) and civic participation in adulthood (Chan etal., 2014; Jennings & Stoker, 2004). Studies have shown that adolescents who engage in sociopolitical debates with family and friends develop a higher sense of belonging (Anderson, 2009; Duke etal., 2009; Talò etal., * Concepción Moreno-Maldonado [email protected] Vanesa Salado [email protected] Tania Gaspar tania.gaspar.bar[email protected] Margarida Gaspar de Matos [email protected] Francisco Rivera [email protected] 1 Department ofExperimental Psychology, Universidad de Sevilla, C/Camilo José Cela, S/N, 41018Seville, Spain 2 Lusófona University, HEI-LAB Portugal, Lisbon, Portugal 3 Department ofDevelopmental andEducational Psychology, Universidad de Sevilla, C/Camilo José Cela, S/N, 41018Seville, Spain 4 Environmental Health Institute, Universidade de Lisboa, ISAMB/Lisbon University, Lisbon, Portugal 9391Current Psychology (2024) 43:9390–9403 1 3 2014), increased social responsibility, and high self-efficacy (Richardson, 2003; Smetana etal., 2006). However, recent studies have voiced concern about low levels of youth social and political participation (Ribeiro etal., 2015), In light of this, our principal research objective is to explore youth engagement in communication about political and social issues (discussions, information seeking) and news media use, and examine its predictive capacity on other important dimensions of adolescent development: sense of unity and perceived social competence (Boyd etal., 2011; Heiss etal., 2020). In addition, both inter- and intra-country analyses are conducted in two neighboring countries: Spain and Portugal. National research in both countries has found relationships between indicators of adolescent civic engagement and sense of belonging or sense of unity (Ribeiro etal., 2015; Salado etal., 2021), in addition to other variables related to social skills (Branquinho etal., 2022). Sociopolitical communication withparents andfriends andinformation seeking asindicators ofadolescent civic engagement Civic engagement is considered to be a complex and multifaceted phenomenon, containing certain problems in its conceptualization (Sherrod etal., 2010). There is a lack of consensus regarding what defines a civically engaged adolescent. Barret and Pachi (2019) offer two concepts to define civic engagement: civic participation (participative behaviors and actions benefitting the community) and psychological commitment (interests, opinions, values, and beliefs regarding political or social issues). Ekman and Amnå (2012) also differentiate between civic engagement –when youth are involved in individual or collective activities outside the political sphere– and political participation –such as voting, protesting, affiliation to political parties, etc. The present research adopts the definition by Schultz etal. (2010), and more specifically one of their indicators: student engagement in communication about political and social issues, which includes debates about sociopolitical issues with family and friends and seeking out information about these issues in newspapers, on the internet, or through national or international televised news. Student self-beliefs (interest, internal political efficacy, and citizenship self-efficacy): indicative of psychological involvement; Student engagement in communication about political and social issues (discussions, information seeking): indicative of individual civic engagement; Student participation in civic activities outside of school: reflects student involvement in collective civic engagement that is not part of the formal learning context; Student participation in civic activities at their schools: reflects student involvement in collective civic engagement that is related to education; and Students’ expected political participation in the future: refers to behavioral intentions with regard to legal and illegal forms as well as individual (electoral) or collective (active political) forms of formal participation (p.115). Adolescents’ political interest is influenced by cognitive changes as well as new social experiences (Neundorf etal., 2013; Russo & Stattin, 2017). Likewise, social interest is the basis of future action and participation in society (Verba etal., 1995; Wanders etal., 2021). Awareness about social problems, discussed within the nuclear family, with friends, or at school (Wanders etal., 2021) fosters the adolescents’ sociopolitical interests (Shehata & Amnå, 2019). For example, research by Lee etal. (2012) demonstrated that youth who discuss public issues with peers and family members, as we as read and reflect upon news, have more interest in and knowledge about sociopolitical problems. Furthermore, communication skills are important for effective civic participation (Battistoni, 2000). An adolescent’s communication skills –developed in interpersonal communication with other people through debating sociopolitical issues– constitute the foundation of their civic competence (McLeod etal., 2010; Shah etal., 2009). Accordingly, youth whose parents are interested in sociopolitical issues will have higher levels of civic knowledge (Schulz etal., 2010), and those who frequently discuss politics with a family member will have a higher probability of volunteering or voting as adults (Zukin etal., 2006). Communicating with parents about sociopolitical issues has been demonstrated to mediate between consumption of sociopolitical information and civic engagement (Boyd etal., 2011). Likewise, adolescents who discuss social problems with their classmates –expressing and listening to different opinions– have shown higher levels of civic knowledge and a higher probability of voting in the future (Torney-Purta etal., 2001). Furthermore, discussing political issues with friends usually contributes to the youth’s interest in politics (Dostie- Goulet, 2009). Similarly, López etal. (2006) observed that 15–25-year-old North Americans who regularly followed the news –either on television or through other means– had a higher probability of participating in political groups, voting, or addressing community problems. In addition, according to the IEA Civic Education Study conducted in 29 countries in 1999 and 2000 (Amadeo etal., 2004), expected adult voter participation in national elections was higher among 14–17-year-olds who frequently read the newspaper compared to those who did not. Accordingly, awareness and understanding the aforementioned issues leads to long-lasting civic attitudes, values, and behaviors during adulthood, contributing to the development 9392 Current Psychology (2024) 43:9390–9403 1 3 of prosocial reasoning (Metzger & Smetana, 2010). Political socialization in different contexts throughout the youth’s formative years will foster civic awareness and the development of their first political and social ideas (Neundorf & Smeets, 2017). Along the same lines, Bandura (1986) highlighted the plasticity of the developmental process, citing systemic and probabilistic relationships between a developing organism and a changing context. The author encourages theories on personality development to consider an adolescent’s social network and the changes and interactions between its members in order to better understand phenomena such as personality, cognition, or behavior. According to Hayden (2006), behaviors have an influence on and are influenced by value systems, social structures, governmental regulations, etc. Thus, adolescent civic engagement, in its different forms, is reinforced by developmental contexts and experiences (Eckstein etal., 2012; Kerestes etal., 2004; Shehata & Amnå, 2019; Stattin etal., 2023). Consequently, following the theory of social capital, the social connections and trust amongst people produced through these experiences will benefit both the individual as well as society (Field, 2008). Therefore, social contexts exert an important influence on an adolescent’s values and civic behavior (Silke etal., 2020). Moreover, given the demonstrated relevance of news media use and discussions about sociopolitical issues as indicators of active and engaged citizens, it leads us to explore whether this indicator can predict other variables of adolescent development such as sense of unity and social self-efficacy. Adolescent engagement incommunication aboutpolitical andsocial issues (discussions, information seeking) asapredictor ofsocial self‑efficacy andsense ofunity As previously mentioned, indicators of civic engagement are related to aspects of adolescent development such as sense of belonging (Anderson, 2009; Duke etal., 2009; Talò etal., 2014) or self-efficacy (Richardson, 2003; Smetana etal., 2006). Likewise, positive development models consider the family, school, or friends to be assets that promote skills for healthy youth development (Benson etal., 2004; Oliva etal., 2010; Scales etal., 2000). This data highlights the importance of considering adolescent social and political communication in their contexts, as well as seeking out sociopolitical information, as predictors of variables such as social self-efficacy and sense of unity. Civic self-efficacy is considered to be a predictor of adolescent participation in different types of civic activity in the future (Ainley & Schulz, 2011; Schulz etal., 2010), and political self-efficacy is associated with democratic school environments which are open to dialogue (Godfrey & Grayman, 2014; Levy, 2013; Manganelli etal., 2015). However, research conducted to date has not yet clarified the relationship between social self-efficacy, as a specific subdimension of the general self-efficacy scale proposed by Bandura etal. (1999) –understood as an individual’s perception of their ability to develop healthy and supportive social relationships (Schunk & Pajares, 2009)– and student engagement in communication about political and social issues. In response to the lack of research on the relationship between these two variables, the authors believe that there could be a positive relationship between news media use and political discussions and social self-efficacy. Similar to adolescents’ political self-efficacy (believing in one’s ability to influence political change)-developed through vicarious experiences and knowledge gained through conversation about political and social issues (Bandura, 2006), it is likely that political debate fosters an individual’s perception of their ability to express opinions, relate appropriately with others, etc. Accordingly, the specific characteristics of the debate could foster a positive attitude towards one’s own ability to participate (Zorn etal., 2006). In addition, certain aspects measured in social self-efficacy related to communication with others are also present in communicative self-efficacy, the latter understood as an individual’s beliefs about their ability to express themselves using their own words, thoughts, and feelings (Koesten etal., 2002). For that matter, various studies confirm that discussing controversial issues can improve communicative self-efficacy (Klosterman & Sadler, 2010; Venville & Dawson, 2010). Sense of unity is conceptualized as the positive feeling derived from forming part of a larger social structure that provides a common good (Larson, 2006; Samdal etal., 2016). Indicators such as seeking information or speaking about sociopolitical issues with family and friends can predict a sense of interdependence with others (Anderson, 2009; Duke etal., 2009; Talò etal., 2014). For example, an open style of communication with the family fosters the youth’s connection to their family unit, and consequently, their connection with others (Boyd etal., 2011). Our research addresses sense of unity as a broader and psychological concept, examining beyond the connection and interdependence of specific contexts. According to the theory of social capital, youth with a stronger connection to and trust of others will develop more social responsibility as a means of maintaining their network of social connections (Wray-Lake & Syvertsen, 2011). In addition to considering the independent associations of these variables with student engagement in communication about political and social issues, we believe it is important to explore the association between this indicator of civic engagement and social self-efficacy through sense of unity, given the current relevance of the latter in other research (Salado etal., 2021; Wahlström etal., 2021). Following 9393Current Psychology (2024) 43:9390–9403 1 3 Smetana etal. (2006), adolescents who participate socially increase their interdependence with others, social responsibility, and the perception of self-efficacy. Lastly, we would like to explore this complex network of relationships according to country of origin. The role ofcountry instudent engagement incommunication aboutpolitical andsocial issues, sense ofunity, social self‑efficacy, andtheassociations betweenthem Both Portugal and Spain foster sociopolitical participation in and out of school. For example, the Dream Teens project in Portugal aims to increase adolescents’ social participation through a participative methodology of research-action (http:// avent uraso cial. com/). In Spain, the Barómetro Opina (Opinion Barometer) (https:// www. barom etroopina. es/) aims to understand youths’ concerns regarding national or international social and political issues. Furthermore, the Ciudades Amigas de la Infancia (Child-friendly Cities) program –sponsored by UNICEF (https:// child frien dlyci ties. org/)– has been implemented in both Spain and Portugal, aiming to design effective policies in any city, town, community, or local government agency, to foster youth social participation and respect for their rights. These projects and activities are indicative of these countries’ consideration of the youth’s opinions in order to improve their wellbeing and foster positive youth development (Baya, 2020), as well as promote different forms of adolescent civic engagement (Adler & Goggin, 2005). For example, youth participants of Dream Teams report a positive impact in their personal development, actions, relationships, and feeling of support (Branquinho etal., 2019). Thus, it can be hypothesized that implementing these types of programs in both countries can help increase different forms of participation, such as that proposed in this study: seek sociopolitical information and participate in debates or discussions with family and friends regarding political and social issues, and simultaneously foster characteristics such as sense of unity and social self-efficacy. Torney-Purta (2002) summarized the analyses conducted in a study with 28 European countries about 14-year-old students’ knowledge and civic skills. In this summary, Torney-Purta (2002) affirms that adolescents from Southern European countries show a high sense of self-efficacy and were considered capable of collaborating with their peers to resolve problems. Schools in these countries promoted student councils, ensured an open-classroom environment for discussing problems, and considered voting and elections to be important. Likewise, perceiving similarity with others and feeling a sense of belonging to a stable and reliable society (Sarason, 1974) may serve as a catalyst for youth social participation (Cicognani & Zani, 2009). A Portuguese focus group study attributed low adolescent participation in civic and political issues to feeling generally excluded by adults, despite parents and media providing information and making them aware of social and political issues (Ribeiro etal., 2015). Accordingly, in a study conducted by Sloam (2016) comparing fifteen European countries it was found that Portuguese adolescents were less likely than their Spanish peers (as well as all other countries) to engage in active forms of social participation (such as strikes or protest, boycotts, etc.). The main aim of this research is to analyze the influence of Spanish and Portuguese adolescent’s participation in communication about sociopolitical issues on social selfefficacy, both directly as well as indirectly through the sense of unity, considering the country of origin. Each scale was examined according to sex, age, and country of origin. It was first hypothesized that there would be differences in the mean scores according to sex and age. The authors believe that boys will have higher scores than girls in civic engagement, but lower in sense of unity and social self-efficacy. In addition, it is hypothesized that age could be key to the development of these indicators, i.e., more civic engagement, sense of unity, and social self-efficacy as age increases. Likewise, the authors argue that the adolescents’ dialogue at home or with peers will increase connections between members, generating a sense of unity which will in turn foster social self-efficacy. Likewise, it is believed that the quality of social and political communication will impact the adolescents’ participation and will be generalized as a phenomenon for developing social and psychological instruments such as social self-efficacy and sense of unity. Lastly, given the similarities between the countries explored, the authors hypothesize that there will be no significant differences in the relationships established in the proposed model. Method Study design andparticipants Population data comes from the 2017/18 edition of the Health Behaviour School-aged Children (HBSC) study, a WHO collaborative cross-national study which has collected data from 11–18-year-old adolescents every 4years since 1983/1984. This research includes 36,992 adolescents 13–18- years old from Portugal (15.4%) and Spain (84.6%), with 18,715 (50.6%) girls and 18,277 (49.4%) boys (Matos & Equipa Aventura Social, 2018; Moreno etal., 2019). Participants were selected using stratified random multistage sampling by conglomerates and data was collected through an online questionnaire, answered anonymously by the adolescents during school hours. Data collection was carried out according to the standardized international protocol (Inchley etal., 2018) and ethical consent was granted by the competent institution of each country: the Comité de Ética de Andalucía (Andalusian 9394 Current Psychology (2024) 43:9390–9403 1 3 Bioethics Committee) in Spain, and the X Comissão de Ética da Faculdade de Medicinada Universidade de Lisboa (10th Ethics Committee of the Lisbon School of Medicine) in Portugal. Instruments The instrument used was the Health Behaviour in Schoolaged Children (HBSC) questionnaire (Inchley etal., 2018). This instrument has three types of questions: mandatory questions (included in all countries), optional questions (specific for different areas), and national questions according to each country’s specific interests. For the present research, the following variables were selected from the optional "Positive health" optional package (Samdal etal., 2017), in addition to sex, age, and country: – Student engagement in communication about political and social issues (SECPS) is one of the indicators used to define civic engagement proposed by Schulz etal. (2010) and measured through 7 items that asked, “How often are you involved in each of the following activities outside of school?”, with four responses options (Never or hardly ever; monthly; weekly; and daily or almost daily). Some of the items were: “talking with your parent(s) about political or social issues”, “watching television to inform yourself about national and international news”, “talking with friends about political and social issues”, etc. Flanagan etal. (2007) uses the 7 items on three different scales, calculating the reliability coefficient for each of them. Thus, item 1 is part of a scale called "Communication with parents about politics", with an alpha coefficient of .86; item 4 is part of the "Communication with friends about politics" scale with an internal consistency of .86; and items 2, 3 and 5 are part of the "Overall media consumption" scale along with other items, with a Cronbach alpha of .81. For items 6 and 7, no validation studies have been found. The present study combines all the items on a single scale with an internal consistency of .85. – Social self-efficacy (SSE) is an 8-item subscale inspired by the self-efficacy scale developed by Muris (2001). Some of the items were “How well can you express your opinions when other classmates disagree with you?”, “How well can you tell other children that they are doing something that you don’t like?”, “How well do you succeed in preventing quarrels with other children?”, etc. The response options oscillated between 1 (not at all) and 5 (very well). The Cronbach alpha of the original sub-scale was .85 (Muris, 2001). In this study, the Cronbach alpha was .86. – Sense of unity (SU) was evaluated through an 8-item measure developed by SALUD project (Stimulating Adolescent Life Skills Through Unity and Drive) and included in the HBSC study (Samdal etal., 2017). Some of the items were: “I feel that I contribute without expecting anything in return”, “I feel that I mean a lot to others”, “I feel that others mean a lot to me”, etc. This scale has been used in other research with alpha values between .76 and .89 (Salado etal., 2021; Samdal etal., 2016; Wahlström etal., 2021). The Cronbach alpha coefficient for this study was .90, showing adequate internal consistency. Data analysis Firstly, descriptive statistics were used for the three variables, including the mean and standard deviation. Secondly, Student t test, ANOVA test and Tukey HSD test was employed to examine mean differences in the examined variables across sex, age and country of origin. Effect size was estimated using Cohen’s d, considering values around .40 as small effect, values between .50 to .70 as intermediate effect, and values equal to or higher than .80 as strong or large effect (Cohen, 1988). In addition, Pearson’s correlations were calculated to examine the associations between the studied variables. Thirdly, the influence of SECPS on SSE through SU was analyzed in a structural equation model with unweighted least squares (ULS). Model fit was evaluated using different adjustment indices: Chi-square (χ2), –its value should not be significant, although Cheung and Rensvold (2002) argue that sample size can influence its value–; Comparative Fit Index (CFI), with values above .90 considered acceptable; Root Mean Square Error of Approximation (RMSEA); and Standardized Root Mean Squared Residual (SRMR). For RMSA and SRMR values near or below .08 and .05 are considered indicators of acceptable model fit, respectively. The intensity of the direct and indirect effects were examined considering the standardized coefficients and their effect sizes. Following Peterson and Brown (2005), the standardized β coefficients were transformed into r and this indicator into eta-square (η2), following the procedure of Dunlap (1994). The etasquare values were interpreted as follows: small effect for values around .05, moderate effect for values from 0.06 to 0.11, and a large effect when the values were equal to or greater than .14 (Cohen, 1988). Lastly, configurational invariance analysis was performed by country of origin and without parameter restrictions. An increase of .01 in CFI was considered to be an indicator of significant change in the model by country (Cheung & Rensvold, 2002). In addition, Fisher’s Z test was employed to examine significant differences in the indicators by country. Only significant differences are reported. IBM SPSS Statistics 22 was employed for descriptive statistics, correlations and mean comparison, and JASP 0.14.1 based on R, was used for the model of structural equations and invariance analysis. 9395Current Psychology (2024) 43:9390–9403 1 3 Table 1 Descriptives, mean comparisons between Spain and Portugal and measure of effect size Descriptive statistics Significance tests and Effect size Total Spain Portugal x SD x SD x SD SECPS Total 14.50 5.30 14.25 5.10 15.60 5.95 t (21447) = - 14.65, p <.001; d = .25 Sex Boys 14.71 5.47 14.56 5.34 15.38 6.01 F (3,21445) = 2607.80; p < .001; d = .23 Tukey HSD test Girls 14.31 5.12 13.94 4.84 15.77 5.89 Spanish boys – Spanish girls: ptukey < .001; d = .12 Spanish boys – Portuguese boys: ptukey < .001; d = .15 Spanish boys – Portuguese girls: ptukey < .001; d = .22 Spanish girls – Portuguese boys: ptukey < .001; d = .28 Spanish girls – Portuguese girls: ptukey < .001; d = .36 Portuguese boys – Portuguese girls: ptukey = .092; d = .06 Age F (5,21443) = 3822,42; p < .001; d = .36 Tukey HSD test 13−14 13.75 5.22 13.48 4.91 14.86 6.18 13-14 Spanish – 15-16 Spanish: ptukey < .001; d = .14 13-14 Spanish – 17-18 Spanish: ptukey < .001; d = .39 13-14 Spanish – 13-14 Portuguese: ptukey < .001; d = .26 13-14 Spanish – 15-16 Portuguese: ptukey < .001; d = .39 15−16 14.41 5.17 14.19 5.01 15.50 5.75 13-14 Spanish – 17-18 Portuguese: ptukey < .001; d = .66 15-16 Spanish – 17-18 Spanish: ptukey < .001; d = .25 15-16 Spanish - 13-14 Portuguese: ptukey < .001; d = .13 15-16 Spanish - 15-16 Portuguese: ptukey < .001; d = .25 17−18 15.75 5.38 15.49 5.28 16.81 5.64 15-16 Spanish - 17-18 Portuguese: ptukey < .001; d = .51 17-18 Spanish – 13-14 Portuguese: ptukey < .001; d = .11 17-18 Spanish - 15-16 Portuguese: ptukey > .999; d = .00 17-18 Spanish - 17-18 Portuguese: ptukey < .001; d = .25 13-14 Portuguese -15-16 Portuguese: ptukey = .013; d = .11 13-14 Portuguese –17-18 Portuguese: ptukey < .001; d = .33 15-16 Portuguese -17-18 Portuguese: ptukey < .001; d = .23 SSE Total 29.79 6.17 30.28 5.51 27.24 7.94 t (20689) = 29.24, p = <.001; d = .50 Sex Boys 29.74 6.38 30.34 5.60 26.93 8.68 F (3,20687) = 287.79; p < .001; d = .41 Tukey HSD test Girls 29.84 5.97 30.43 5.43 27.49 7.29 Spanish boys – Spanish girls: ptukey = .733; d = .01 Spanish boys – Portuguese boys: ptukey < .001; d = .54 Spanish boys – Portuguese girls: ptukey < .001; d = .47 Spanish girls – Portuguese boys: ptukey < .001; d = .57 Spanish girls – Portuguese girls: ptukey < .001; d = .50 Portuguese boys – Portuguese girls: ptukey = .025; d = .07 Age 13−14 29.71 6.59 30.53 5.64 26.48 8.75 F (5,20685) = 180.66; p < .001; d = .42 Tukey HSD test 13-14 Spanish – 15-16 Spanish: ptukey = .030; d = .04 13-14 Spanish – 17-18 Spanish: ptukey = .453; d = .04 13-14 Spanish – 13-14 Portuguese: ptukey < .001; d = .63 13-14 Spanish – 15-16 Portuguese: ptukey < .001; d = .49 15−16 29.84 5.95 30.30 5.44 27.56 7.63 13-14 Spanish – 17-18 Portuguese: ptukey < .001; d = .44 15-16 Spanish – 17-18 Spanish: ptukey > .999; d = .00 15-16 Spanish - 13-14 Portuguese: ptukey < .001; d = .61 15-16 Spanish - 15-16 Portuguese: ptukey < .001; d = .46 15-16 Spanish - 17-18 Portuguese: ptukey < .001; d = .41 17-18 Spanish – 13-14 Portuguese: ptukey < .001; d = .59 9396 Current Psychology (2024) 43:9390–9403 1 3 Table 1 (continued) Descriptive statistics Significance tests and Effect size Total Spain Portugal x SD x SD x SD SECPS Total 14.50 5.30 14.25 5.10 15.60 5.95 t (21447) = - 14.65, p <.001; d = .25 17−18 29.84 5.84 30.31 5.44 27.94 6.94 17-18 Spanish - 15-16 Portuguese: ptukey < .001; d = .46 17-18 Spanish - 17-18 Portuguese: ptukey < .001; d = .41 13-14 Portuguese -15-16 Portuguese: ptukey < .001; d = .13 13-14 Portuguese –17-18 Portuguese: ptukey < .001; d = .18 15-16 Portuguese -17-18 Portuguese: ptukey = .656; d = .05 SU Total 28.53 7.08 29.18 6.46 25.71 8.79 t (20830) = 28.13, p = <.001; d = .50 Sex Boys 28.44 7.41 29.18 6.73 24.88 9.26 F (3,20828) = 279.22; p < .001; d = .43 Tukey HSD test Girls 28.63 6.77 29.19 6.18 26.38 8.34 Spanish boys – Spanish girls: ptukey < .999; d = .00 Spanish boys – Portuguese boys: ptukey < .001; d = .59 Spanish boys – Portuguese girls: ptukey < .001; d = .39 Spanish girls – Portuguese boys: ptukey < .001; d = .63 Spanish girls – Portuguese girls: ptukey < .001; d = .42 Portuguese boys – Portuguese girls: ptukey < .001; d = .17 Age 13−14 28.91 7.56 29.84 6.71 25.10 9.43 F (5,20826) = 181.19; p < .001; d = .42 Tukey HSD test 13-14 Spanish – 15-16 Spanish: ptukey < .001; d = .15 13-14 Spanish – 17-18 Spanish: ptukey < .001; d = .17 13-14 Spanish – 13-14 Portuguese: ptukey < .001; d = .65 13-14 Spanish – 15-16 Portuguese: ptukey < .001; d = .55 15−16 28.35 6.87 28.84 6.34 25.97 8.64 13-14 Spanish – 17-18 Portuguese: ptukey = .989; d = .52 15-16 Spanish – 17-18 Spanish: ptukey < .001; d = .01 15-16 Spanish - 13-14 Portuguese: ptukey < .001; d = .53 15-16 Spanish - 15-16 Portuguese: ptukey < .001; d = .42 15-16 Spanish - 17-18 Portuguese: ptukey < .001; d = .39 17-18 Spanish – 13-14 Portuguese: ptukey < .001; d = .51 17−18 28.26 6.64 28.75 9.43 26.28 7.93 17-18 Spanish - 15-16 Portuguese: ptukey < .001; d = .41 17-18 Spanish - 17-18 Portuguese: ptukey < .001; d = .38 13-14 Portuguese -15-16 Portuguese: ptukey = .011; d = .09 13-14 Portuguese –17-18 Portuguese: ptukey < .001; d = .14 15-16 Portuguese -17-18 Portuguese: ptukey = .898; d = .04 Note: Data from the HBSC study in Spain and Portugal (2018 edition). Own elaboration. SECPS, Student engagement in communication about political and social issues; SSE, Social self-efficacy; SU, Sense of unity; x, Means; SD, standard deviation; t, student t; F, Anova test; d, Cohen’s d Results Descriptive statistics andcorrelations Table1 presents the descriptive statistics, including mean and standard deviation values for continuous variables, as well as the results of the means comparisons by sex, age, and country. As shown in Table1, Student's t results showed significant differences across countries. Portuguese adolescents showed higher SECPS scores with small effect size, while Spanish adolescents showed higher SSE and SU with intermediate effect size. Regarding sex differences, boys have higher scores than girls in SECPS in both the global sample and in Spain, however these differences were not observed in Portugal. Concerning SSE and SU, girls scored higher than boys in the global score, but no sex differences were identified in SSE and SU when comparing Spanish boys and girls. Analysis by age showed that as age increases so does SECPS –as well as a slight increase in SSE– both in the global sample and by country. However, there are no significant differences between Spanish 17–18-year-olds and the other two age groups, and between Portuguese 17–18 and 15–16-year-olds. Lastly, SU is higher in 13–14-year- old boys, with significant differences compared to the other groups in both the global sample and in Portugal. However, 17–18-year-old Portuguese youth have higher SU than the 9397Current Psychology (2024) 43:9390–9403 1 3 rest of the groups, despite no significant differences between 15–16 and 17–18-year-old Portuguese youth. Lastly, Pearson’s correlation coefficient values showed that SSE and SU have positive and significant relationships (p < 0.05) with SECPS scores, with a low intensity. In addition, SSE and SU are positively associated, with moderate intensity. Testing amediation model oftheinfluence ofstudent engagement incommunication aboutpolitical andsocial issues onsocial self‑efficacy throughsense ofunity andinvariance acrosscountries As presented in Table2, goodness of fit indices for the hypothesized mediation model showed excellent data fit (CFI = .980; RMSEA = .054; SRMR = .041) despite Chisquare being significant (χ2 = 12,226.710; p < .001). Figure1 presents the standardized coefficients of the global model estimating the direct association between SECPS and SSE, as well as the indirect effects of SECPS on SEE through SU. The model explained 34.2% of the variance of SSE and 6.2% of the variance of SU. The direct path from SECPS to SSE showed low and significant factor loading (β = .12; η2 = .02) Likewise, mediation effects suggested that the effect of SECPS on SEE through SU operates significantly and with low intensity (β = .13; η2 = .03). In both cases, the effect size was small. However, the directs effects of SECPS on SU (β = .25; η2 = .09) and of SU on SEE (β = .54; η2 = .30) showed greater intensity, with intermediate and large effect sizes. Configurational invariance analyses were performed with samples segmented by country. Figure2 presents the standardized coefficients for Spain and Portugal. The models explained 34.1% (for Spain) and 29.0% (for Portugal) of SEE, and 7.2%% (for Spain) and 12.4% (for Portugal) of SU. The configurational invariance model fit was excellent for country (CFI = .979; RMSEA = .059; SRMR = .047) with a .003 decrease in CFI (values must be below .01 in order to consider invariance). Table 2 Goodness-of-fit indices for the proposed factorial model and the configurational invariance analysis Data from the HBSC study in Spain and Portugal (2018 edition). Own elaboration. χ2/ df, aChi-square/degree of freedom; bNNFI, Non- Normed Fit Index; cCFI, Comparative Fit Index; dIFI, Incremental Fit Index; eRMSA, Root Mean Squared Error; fCI, confidence interval; gSRMR, Standardized Root Mean Squared Residual; h ∇ CFI, Decrease in CFI Global Country χ2/dfa53.86 32.63 NNFIb.980 .977 CFIc.982 .979 IFId.982 .979 RMSAe (CI 95%)f.054 .059 SRMSg.041 .047 ∇ CFIh- .003 Fig. 1 Representation of the standardized estimations of the path coefficients of the global model. Note: Data from the HBSC study in Spain and Portugal (2018 edition). Own elaboration. SECPS, Student engagement in communication about political and social issues; SSE, Social self-efficacy; SU, Sense of unity Fig. 2 Standardized solution of the final model across the different groups by country. The data is presented as Spain/Portugal. Note: Data from the HBSC study in Spain and Portugal (2018 edition). Own elaboration. SECPS, Student engagement in communication about political and social issues; SSE, Social self-efficacy; SU, Sense of unity 9398 Current Psychology (2024) 43:9390–9403 1 3 In addition, small variations in the standardized coefficients were examined using Fisher’s Z, showing that changes were not significant (p < .05), except in the association between SECPS and SU which was significantly more intense in Portugal (Z = -6.16, p < .001) and between SU and SEE which was significantly greater in Spain (Z = 6.45, p < .001). However, these results only show relationship strength, without affecting the rest of the indicators that reveal the model’s configurational invariance across the country. Discussion The main objective of this research was to analyze whether adequate socio-political and informative communication influences social self-efficacy –either directly or through sense of unity– promoted by the interactions with others, in a representative sample of adolescents residing in Portugal and Spain. Consistent with previous findings, results showed that establishing debates with family and friends is associated with youth self-efficacy (Richardson, 2003) and promotes a recognized interdependence with others (Talò etal., 2014). Likewise, the influence of sociopolitical debate with family and friends on sense of unity, as observed in this research, has been confirmed by other studies which highlight the importance of political debate at home for reinforcing active civic engagement in the future (Hooghe & Stiers, 2022). Prior to examining the relationships in the proposed model, each of the variables was explored individually and according to different sociodemographic variables. Thus, results showed country differences in the scores of the three variables (communication about politics and social issues with family, friends, and media; social self-efficacy, and sense of unity). Regarding social media and political communication, Portuguese adolescents scored higher in engagement in communication about political and social issues than their Spanish peers. These results contrast with other research. For example, Magalhães and Moral (2008) characterized Portuguese youth as having low exposure to political information and therefore lower levels of political knowledge due to their notorious discontent with the functioning of their democratic system. In another study, Sloam (2016) compared five forms of political participation –voting in national elections, displaying a badge, signing a petition, joining a boycott, and participating in a demonstration– across fifteen European countries. Results showed Portuguese adolescents participating at comparatively lower levels than Spaniards, as well as presenting lower participation than adolescents from the other countries such as Sweden and France or those that are part of continental European democracies such as Germany, Belgium, Austria, and Luxembourg. In this sense, and considering our results, it is possible that countries ‘specialize’ in certain forms of participation according to their cultural or social characteristics, with Portuguese youth exercising more passive forms of socio-political participation such as social and political debates and searching for information. This coincides with Menezes etal. (2005), who found in the International Study of Civic Education (Estudo Internacional sobre Educação Cívica) that Portuguese youth –compared to other European youth– engaged more in passive-conventional or social activities such as the tendency to discuss politics mainly with parents and very rarely with teachers. Secondly, and regarding social self-efficacy and sense of unity, Spanish adolescents showed higher scores than their Portuguese peers. These results must be interpreted with caution due to the lack of prior studies comparing the two countries in this subdimension of general self-efficacy by Bandura etal. (1999), nor with the new dimension proposed by Samdal etal. (2016): sense of unity. In fact, previous research conducted by Verdugo etal. (2018) using the general self-efficacy scale of Bandura etal. (1999) have found no differences between Spanish and Portuguese adolescents. Likewise, research about the social skills of native and immigrant adolescents from Spain and Portugal found high mean scores in both groups and in both countries (Romero-Oliva etal., 2017). On the other hand, Spain and Portugal are considered to be two very culturally similar countries (Instituto de Política Familiar, 2014; Inchley etal., 2016; Pommier etal., 2010), in which their citizens function interdependently and synchronously in relation to other members (Rokach & Nieto, 2000). Therefore, as the 2012 Report on Youth in Spain (Moreno & Rodríguez 2013) points out, we can hypothesize that Spanish youth tend to have a more-positive self-evaluation than their peers in other European countries. Analyses by country, sex, and age produced interesting results. When asked if they participated in sociopolitical discussions with family and friends, or if they would seek out sociopolitical information in the media, Spanish boys scored higher than girls. This data coincides with prior research on civic engagement and gender, reporting that boys are more interested in following news or have a higher intention of voting in the future (Wray-Lake etal., 2020) and that girls are more interested in volunteering (Gaby, 2017). Sex differences in sense of unity and social self-efficacy were also found in favor of girls, except when comparing Spanish youth. Research focusing on these variables have shown discrepancies. It seems that girls have higher social self-efficacy than boys at a younger age (Coleman, 2003), however these differences dissipate as they get older (Bacchini & Magliulo, 2003; Gaudiano & Herbert, 2007). Little research has been conducted about sense of unity, as it is a relatively new construct. However, a study by Wahlström etal. (2021) found no sex differences in a sample of Swedish adolescents.