1.1 Social interaction skills

Social interaction skills are a group of observable actions in the daily life context. They are learned, socially acceptable behaviors that enable effective interactions with others (Fisher & Marterella, 2019; Grover et al., 2020). Social interactions are essential for children and adolescents to achieve developmental milestones and learn how to integrate into society (Sezici et al., 2017). Most adolescents have an increased need for social interaction, including peer acceptance and approval (Foulkes & Blakemore, 2016; Sawyer et al., 2018; Somerville, 2013). High-quality social interactions with peers appear to protect against mental health problems (e.g., stress and depression) and strengthen adolescent resilience and well-being (Platt et al., 2013; Van Harmelen et al., 2017). Moreover, people with acquired SIS can more easily change and adapt to circumstances and have higher life satisfaction and happiness (Hall & Merolla, 2020; Holt-Lunstad, 2021; Quoidbach et al., 2019). In addition, good SIS are associated with lower loneliness (Piquart & Sorensen, 2003) and mortality (Holt-Lunstad, 2021) rates.

The International Classification of Functioning, Disability, and Health (ICF; World Health Organization, 2001) describes and organizes information on functioning and disability. The ICF defines participation as active engagement in life situations that influence health and well-being. The ICF child and youth version (ICF-CY; World Health Organization, 2007) provides a comprehensive inventory of various aspects of child and adolescent functioning and offers a holistic perspective on the barriers and facilitators influencing participation in life situations, including social activities. It standardized the language and conceptual basis to define and measure disability and health.

The ICF-CY contains two main components. The first, body functions and structures, refers to the constructs of change in physiological functions or anatomical structures in body function and structure domains. The second, activities and participation, describes activity as the “execution of a task or action by an individual,” and participation as “involvement in a life situation” (World Health Organization, 2007). It categorizes SISs as basic or complex. Basic interpersonal skills describe interacting with people in a contextually and socially appropriate manner, such as showing consideration and esteem when appropriate. Complex interactions relate to maintaining and managing those interactions in a contextually and socially appropriate manner (e.g., regulating emotions and impulses and controlling verbal and physical aggression).

The ICF-CY lists skills under each category, but the list is cumbersome for assessing these skills. This study uses the ICF-CY model as a conceptual basis to reference common SIS among adolescents. Associating items in our suggested assessment tool for measuring SIS with the corresponding ICF-CY codes helped determine whether the instrument encompasses a comprehensive view and most aspects outlined in this framework.

1.2 Social media interactions

According to prior studies, adolescents worldwide perceive social media environments as integral to their education, culture, and social lives (Keles et al., 2020; O'Reilly et al., 2018), and most use them for several hours daily (Chang et al., 2022; Woods & Scott, 2016). This situation has raised inquiries regarding the influence and consequences of social media on the social skills of present and upcoming generations (Blakemore & Agllias, 2020). Social media is any digital tool or application that allows users to interact socially (Strasburger et al., 2013). Users of these environments can generate and share content with others, create personalized profiles, and develop online social networks (Obar & Wildman, 2015). It has been argued that using technology potentially enables the continuity of relationship tasks. However, whether it can address all the adolescents' social needs has not been clearly answered (Orben et al., 2020). Although the value of social media use varies among individuals, social people are active both online and offline (Stepanikova et al., 2010). Some people benefit from social media's opportunities to increase collaboration and tolerance (Guinta & John, 2018), whereas others experience more negative effects (e.g., Hanley et al., 2019; Pouwels et al., 2022).

A review of adolescents' social-media-use patterns from 2006 to 2016 in the United States suggested that their use was extremely high (Twenge et al., 2018). In a U.S. representative survey, 77% of adolescents reported that social media was important for keeping up with friends daily, and 69% reported it as important for having meaningful conversations with close friends (Clark et al., 2018). These patterns of increasing social media use intensified with COVID-19 social-distancing restrictions (World Health Organization, 2021).

1.3 In-person social interactions

Adolescents are increasingly the time they spend on social media. Thus, a growing body of research has examined associations between adolescents' social patterns and skills in social media communications versus the quality of their in-person social interactions (Nangle et al., 2020). For example, Spies Shapiro and Margolin's (2014) review showed that adolescents with strong in-person SIS tended to have more social connections on social media. Assessing adolescents' SIS in person and on social media is complex (Cavell, 1990). Considering the wide range of skills social functioning encompasses, some measures aim to include those diverse domains broadly, whereas others focus on assessing a specific skill subset (Crowe et al., 2011). A wide variety of techniques, including rating scales, observations in unstructured and structured contexts, interviews, self-reports, and sociometric assessments, can be used to assess social competence. Each technique has advantages and disadvantages; hence, it has been recommended that SIS evaluations be collected from multiple informants and contexts (Nangle et al., 2020). Because adolescents can access a range of social interactions according to the various contexts in which they are involved, self-report assessments might provide the best insights into their social competence (Junttila et al., 2006).

To our knowledge, only one measurement tool has been developed to evaluate several social skills dimensions in two contexts (i.e., face-to-face and computer-mediated communication). According to its authors, the Real and Electronic Communication Skills Questionnaire (RECS; Mantzouranis et al., 2019) is a useful and promising instrument to capture social skills in online and offline contexts. This tool consisted of a 100-item questionnaire comprising two subscales (face-to-face and online contexts).

Our study also suggests a self-report questionnaire for adolescents to review their social skills in two contexts: in-person and online (social media). Unlike the RECS, this proposed tool includes a single list of skills, based on the ICF-CY list, suitable for both environments. Reich's (2017) literature review found no studies that investigated social competence in online spaces or identified links between online and offline social competence. Thus, our suggested questionnaire steps the research in this area forward by allowing a one-to-one comparison of social skills in both environments.

This study had three objectives: (1) to describe the development and psychometric properties of the Social Interaction Skills Questionnaire (SISQ) designed to help fill the research and clinical gap and spotlight the adolescents' viewpoint regarding their SIS in person versus on social media; (2) to assess differences in adolescents' perspectives on their SIS abilities in person versus on social media; and (3) to assess the relationships between SIS and subjective well-being through correlations and prediction.

2.1 Participants

Following G*Power calculated for detecting medium effects with a statistical power of 95%, alpha of 0.05, and up to 15 predictors (Buchner et al., 2013), the study included 214 adolescents with typical development aged 12 through 18 years (M = 15.3 years, SD = 1.77); 131 girls (61.2%) and 83 boys (38.8%). Of them, 55.1% (n = 118) lived in urban areas and 44.9% (n = 96) in rural settings. Twenty (9.3%) participants were in Grade 7, 35 (16.4%) in Grade 8, 33 (15.4%) in Grade 9, 36 (16.8%) in Grade 10, 48 (22.4%) in Grade 11, and 42 (19.6%) in Grade 12. Inclusion criteria required the participants to be attending regular (mainstream) schools and have self-reported typical development (i.e., no known developmental deficits, such as psychiatric or emotional disorders, autistic spectrum disorders, physical disabilities, or neurological diseases).

2.2 Procedure

After receiving the University of Haifa's Ethics Committee approval (253/21, 24 March 2021), we published an advertisement on social media inviting adolescents to participate in the study. The advertisement included a link to the research description, its importance, and how to participate. It also clarified that the answers would be completely anonymous, with no way to identify the participants. Adolescents who met the inclusion criteria, agreed to participate, and signed online consent forms were referred to the research questionnaire. The data were collected after the 2021 COVID-19 lockdown period (June–October 2021), when the education system had returned to full and routine operation alongside the COVID-19 precautions.

2.3 Instruments

We designed an online questionnaire containing all the instruments described in the following sections.

2.4 Data analysis

Data analysis was conducted using IBM SPSS (Ver. 27). We calculated ranges, means, and standard deviations for descriptive statistics. Content analysis was based on the COSMIN model suggestions (Terwee et al., 2012) using focus groups. For construct validity, a Kaiser-Meyer-Olkin (KMO) measure of sampling adequacy was used to assess the respondent data suitability for factor analysis (Kaiser & Rice, 1974). The KMO index ranges from 0 to 1, with 0.65 considered suitable for factor analysis. In this study, the KMO index was 0.90, and the Bartlett test of sphericity yielded p < 0.001, indicating suitability for factor analysis. Exploratory factor analysis (EFA) with principal component extraction and oblique rotation was performed to examine the SISQ's underlying construct validity (Altman, 1991). Items that yielded factor loading values of less than 0.35 were eliminated from the pool. Cronbach's alpha was calculated for internal reliability, and t tests were conducted to assess differences in SIS, with Cohen's d calculation for effect sizes. We also conducted Pearson correlations and hierarchical regression analysis to examine correlations between SIS and well-being.

3.1 SISQ psychometric properties

Following the EFA, all 19 original items were extracted into a monofactorial component. Table 1 presents the items' factor loadings and factor percentages of variance. The EFA was performed on the in-person and social media interaction skills items to confirm adequacy. Following the factor analysis, Items 7 and 14 were omitted from the analysis of each environment; thus, the final questionnaire was left with 17 items. Cronbach's alpha was calculated for the remaining 17 items, indicating high internal reliability for in-person (α = 0.90) and social media (α = 0.90) SIS.

3.2 Differences between in-person and social media SIS

In total, the adolescents rated their in-person SIS (M = 3.98, SD = 0.76) significantly higher than their social media SIS (M = 3.76, SD = 0.77), t(213) = 5.49, p < 0.001, d = 0.38. We conducted an item-level analysis to identify specific skills rated differently between the environments. For this purpose, a paired t test was performed for all 17 items with Bonferroni correction (p < 0.002 considered significant), and Cohen's d effect size was calculated. In seven items, a significant difference was found in the participants' ability to express their SIS; thus, they rated their ability in the in-person environment higher in all those items (Table 2).

In assessing the difference between the two importance-perception questions (participating and engaging in the in-person vs. the social media environment), they reported that participating and being engaged in the in-person environment (M = 4.29, SD = 0.92) was more important to them than participating and being engaged in the social media environment (M = 3.43, SD = 1.10), t(213) = −11.57, p < 0.001, d = 0.79.

3.3 Correlations between SIS and well-being

The adolescents' well-being scores ranged between 20% and 100%, with a mean score of 62.83% (SD = 15.70). No significant correlations were found between age and well-being, and no gender differences were found in well-being. A significant correlation was found between the adolescents' subjective well-being and both in-person SIS (r = 0.40, p < 0.001) and social media SIS (r = 0.36, p < 0.01). Table 3 presents the correlations between the variables. Fisher's Z comparing those two correlations with well-being indicated no significant differences (Z = −0.60, p = NS).

A hierarchical regression analysis examined whether the SISQ variables (total in-person and social media scores) related to the adolescents' reported well-being. Age was included in the model as a control independent variable in the first step because age correlated with in-person SIS (but age was not related to well-being). Only in-person SIS predicted 18% of the explained variance (Table 4).

4.1 SISQ psychometric properties

We found the SISQ to be highly reliable. Its content validity was obtained from the literature, representatives of the relevant populations, and experts following the guidelines in the literature (Yaghmaie, 2003). The EFA revealed that all 17 final SISQ items loaded onto a single factor, supporting the questionnaire's monofactorial structure. This finding suggests that the SISQ effectively captures the underlying construct of SIS in both the in-person and social media environments. Further, the SISQ's high internal reliability, as indicated by Cronbach's alpha coefficients for in-person and social media SIS, demonstrates that the items within each environment consistently measure the intended construct. This result suggests that the SISQ items are reliable indicators of SIS in in-person and social media contexts.

4.2 Conclusions and recommendations for future research

This study suggests a new questionnaire (SISQ) for assessing adolescents' SIS in person and on social media. The SISQ was found to be a valid, reliable tool to assess adolescents' SIS in both environments.

Although social media is a leading environment in which adolescents manage their social life and relationships with peers, our results show that the adolescents reported fewer SIS in this environment compared to in-person interactions. To date, interventions to promote social media SIS were developed mainly for adolescents with disabilities. For example, Raghavendra et al. (2013) investigated the effectiveness of tailored, one-on-one support strategies to facilitate the social participation of youth with disabilities by using the Internet for social networking. A more recent study showed a postintervention increase in participants' performance on social media problem areas and their satisfaction with that performance. Moreover, a significant increase in the number of online communication partners was found. Interviews with the adolescent participants highlighted increased social participation, independence, and literacy (Raghavendra et al., 2018).

Our study's results should be interpreted cautiously primarily because of its relatively small sample. We recommend further research among larger adolescent populations with typical development and with developmental disabilities as a group with high vulnerability to limited social interactions. In addition, we collected our data in a country with a relatively homogenous culture. Further worldwide research among more diverse populations would yield broader conclusions. Finally, despite the presented correlations with well-being, the current cross-sectional study design does not indicate the directionality of those correlations.