Predictive Relationships Between Social Anxiety, Internet Addiction and Alexithymia in Adolescents

The purpose of this study is to determine whether social anxiety predicts internet addiction and alexithymia in adolescents. The study group of the study consists of 406 adolescents (216 women, 190 men), with the following three age groups: 13–14 (32.3%), 15–16 (43.8%), and 17–18 (23.9%). “Social Anxiety Scale for Adolescents (ESCO)”, “Internet Addiction Scale”, “Toronto Alexithymia Scale” and “Personal Information Form” were used as data collection tools. Structural equation modeling was made in the analysis of the data. According to the findings, there was a positive linear relationship between social anxiety and internet addiction. When the predictive relationships between social anxiety and alexithymia were examined, it was determined that there was a positive linear relationship. The findings were discussed in the light of the relevant literature.

Individuals with alexithymia cannot define their emotions and cannot determine the causes of their emotional responses (Lumley, 2004;Speranza, Loas, Wallier, & Corcos, 2007), they lack imagination and may show behaviorally attenuated effects (Campanella et al., 2014;Vorst & Bermond, 2001). In addition, patients with alexithymia show high alexithymia and low social support characteristics (Geenen, van Ooijen-van der Linden, Lumley, Bijlsma, & van Middendorp, 2012). Since Alexithmic individuals are cold and distant, the lack of interpersonal relationships can cause problems in their social lives (Goerlich-Dobre, Witteman, Schiller, van Heuven, Aleman, & Martens, 2014;Vanheule, Desmet, & Meganck, 2007). Alexithymic individuals show emotional awareness and communication deficiencies (Lumley, Neely, & Burger, 2007). Alexithymia consists of four main elements: difficulty in identifying and explaining emotions; limited creative processes; avoiding the tendency to act to resolve contradictions; and difficulty in making a detailed description of facts, events and physical symptoms (Taylor, 2000).

Internet Addiction
The concept of internet addiction was first introduced by Young (1996). Various concepts such as pathological internet use (Davis, 2001), problematic internet use (Davis, Flett, & Besser, 2002) have been used to describe this behavior. The internet is a highly used technological tool and it is difficult to detect addiction. Therefore, it is important to understand the criteria that distinguish normal from pathological internet use (Young, 2004).
Problematic internet use or internet addiction are excessive or poorly controlled concerns, urges, or behaviors that cause trouble with internet use (O'Reilly, 1996;Weinstein & Lejoyeux, 2010). Internet addiction can be conceptualized as an incompatible model of internet use behavior associated with various psychological and social problems (Chou, Condron, & Belland, 2005), the inability of the individual to control the internet use causes sadness and functional disruption of daily activities (Sato, 2006;Shapira et al., 2003). Dependency, according to experts, must harm various levels of function, such as family, social, school, professional and psychological functionality (Gentile, 2009). It also consists of a heterogeneous range of internet activities with potential disease value, such as gaming, shopping, gambling, or social network (Kuss & Griffiths, 2012).
Internet for young people; It offers opportunities such as e-mail, chat, discussion group, social entertainment (games, songs, videos) and so on (Chou & Hsiao, 2000). In addition, the internet has negative effects and these effects can create an isolating environment that leads to loneliness and less social interaction with family members and friends (Turkle, 1996). Adolescents constitute the biggest target group related to the negative effects of internet addiction (Chou, Condron, & Belland, 2005). There are studies indicating that the prevalence of internet addiction among high school students is high (Sasmaz et al., 2013).
One of the reasons why adolescents are internet addicts is that the internet provides the opportunity to access many Web sites and communicate with various friends by meeting personal needs with personal computers (Yang, 2001). Adolescents with internet addiction are more likely to have substance abuse experience. They stated that adolescents with high Internet addiction seek a higher rate of innovation and harm avoidance (Ko, Yen, Chen, Chen, Wu, & Yen, 2006).

Research Question
What is the relationship between social anxiety, internet addiction and alexithymia in adolescents? With the structural equation model. For this purpose, the research was carried out according to the relational screening model, which is a sub-type of the general screening model.

Social Anxiety Scale for Adolescents (SAS-A)
The social anxiety scale for adolescents consists of 22 items, four of which are unrelated. The scale consists of three factors. Supported by factor analysis, these sub-scales are Fear of Negative Evaluation (FNE), General Social Avoidance and Distress (SAD-G) and Social Avoidance and Distress in New Situations (SAD-N). Cronbach Alpha value of the sub-dimensions of the scale was found to be .86 for Fear of Negative Assessment, .82 for the sub-dimension of Fear in New Social Situations, .78 for the sub-dimension of Fear and Restlessness in the General Social Situations, and .95 for the whole scale. The correlation between the sub-dimensions of the scale varies between .61 and 0.64 (p < 0.01) (Aydın & Tekinsav-Sütçü, 2007). DFA analysis was conducted on the social anxiety scale on the study group. As a result of the CFA analysis, it was observed that the factor loads of the scale items ranged from .41 to .83 and each scale item was significant (x 2 = 4.11).

Internet Addiction Scale (IAS)
The scale consists of 35 items. The scale is rated as a five-point Likert type, ranging from 'strongly disagree' (1) to 'strongly agree' (5). As a result of the CFA analysis, it was observed that the factor loads of the scale items ranged from .41 to .83 and each scale item was significant (x 2 = 4.11). The average age of the participants is 15.8. Cronbach Alpha internal consistency coefficient of the scale was found to be 0.94. The scale consists of 4 sub-factors. These are "Withdrawal", "Controlling difficulty", "Disorder in Functionality" and "Social Isolation". The total explained variance for its four sub-dimensions is 47.46%. The high score obtained from the scale indicates internet addiction (Günüç & Kayri, 2010). DFA analysis was carried on the scale of Internet addiction out on the study group. As a result of the CFA analysis, it was observed that the factor loads of the scale items ranged from .58 to .78 and each scale item was significant (x 2 = 3.55).

Toronto Alexithymia Scale (TAS-20)
The original form of the Twenty-question Toronto Alexithymia Scale (TAS-20) was developed by Bagby, Parker and Taylor et al. (1994) and adapted to Turkish by Gulec (2009). The internal reliability coefficient of the scale was calculated as 0.78. And of the sub-scales were between 0.57-0.80. TAÖ-20 is a Likert-type self-rating scale, scored between one (never agree) and five (always agree). Difficulty Identifying Feelings sub-scale consists of seven items and is defined as the difficulty in identifying emotions and distinguishing them from bodily sensations accompanying emotional arousal. Difficulty Describing Feelings sub-scale consists of five items and it is defined as difficulty in transferring emotions to others. Externally-Oriented Thinking sub-scale consists of eight items, and it is defined as the existence of an extroverted cognitive structure, the weakness of introverted thinking and imagination. Individuals are asked to mark the most appropriate one of the options of "Never", "Rarely", "Sometimes", "Often" and "Always" for each item. High scores indicate high alexithymic level (Güleç et al., 2009)

Personal Information Form
It was prepared by the researchers to determine the gender and age of the participants.

Data Collection and Analysis
For the purpose of collecting data; The Social Anxiety Scale for Adolescents (ESCO), "Internet Addiction Scale", "Toronto Alexithymia Scale" and "Personal Information Form" were applied to the participants determined by the appropriate sampling method. A total of 406 students were recruited through convenience sampling method. In the research, structural equation modelling analysis was done in order to test the relationship between the variables of adolescents' social anxiety, internet addiction and alexithymia through the structural equation model. Structural equation modelling analysis was carried out with the AMOS 19 Program. In addition, correlation analysis and frequency, percentage analysis were done with SPSS 20 data analysis program. jel.ccsenet.o

Results
The mode (EOT, DD found stati (TLI) and correlation  Vol. 9, No. 2; When the model in Figure 1 and the data in Table 2 are examined, it is seen that the social anxiety variable affects the alexithymia variable (t = 6.11, p < 0.01). The connection coefficient value for this factor was determined as β = 0.51. When the predictive relationships between social anxiety and alexithymia variables were examined, it was determined that there was a positive linear relationship. In other words, findings reveal that as social anxiety level increases in adolescents, alexithymia will increase. The mean scores, standard deviations, and correlation values between the variables are presented in Table 3. The findings show that social anxiety is significant and in the same direction with internet addiction (r = .23, p < .01); social anxiety is significant and in the same direction with alexithymia (r = .29, p < .01).

Discussion
In this part of the study, the findings obtained by examining the relationship between the social anxiety, internet addiction and alexithymia variables of the adolescents constituting the study group were discussed and discussed. When the predictive relationships of adolescents between social anxiety and internet addiction variables were examined, it was determined that there was a positive linear relationship. In other words, findings reveal that as adolescents increase social anxiety, internet addiction will increase.
Internet usage frequency has a negative effect on frequency of communication with others and social interaction (Nie, 2001). For example; social anxiety has also been shown to increase as adolescents use the internet (Harman, Hansen, Cochran, & Lindsey, 2005). Socially anxious individuals may find it easier to interact online rather than face-to-face interactions where observation by others may have a fear of negative evaluation (Shepherd & Edelmann, 2005). From this point of view, studies supporting the findings obtained in the research have been carried out. Lee and Stapinski (2012) found social anxiety as an important predictor of problematic internet use. In another study, social anxiety predicts problematic internet use (Weinstein & Lejoyeux, 2010). In the model created for social anxiety and internet addiction for six Asian countries, the path from social anxiety to internet addiction; it fits well for Japan, South Korea and the Philippines (Lai et al., 2015). Weinstein, Dorani, Elhadif, Bukovza and Yarmulnik (2015) found a positive relationship between internet addiction and social anxiety in their study. Social anxiety is positively associated with problematic internet use (Cuhadar, 2012;Prizant-Passal, Shechner, & Aderka, 2016). Huan, Ang and Chye (2014)'s results obtained revealed support for a model in which social anxiety fully mediated the relationship between the psychological variables of shyness and loneliness and problematic internet use. Fear-anxiety and depression were positively correlated with cognitive internet use cognitions among university students (Durak & Senol-Durak, 2013). It has been found that students with internet addiction have higher social anxiety (Yucens & Uzer, 2018). Anxiety has been found to positively predict internet addiction (Azher, Khan, Salim, Bilal, Hussain, & Haseeb, 2014;Razieh, Ghasempoor, Ajdari, & Sadeghigooghari, 2012).
High level of social anxiety is also associated with frequent internet use (Mazalin & Moore, 2012). The internet may have good potential as an alternative tool to provide interventions for social anxiety (Yen, Yen, Chen, Wang, Chang, & Ko, 2012) and also increased face-to-face contacts can help reduce symptoms of internet addiction (Yao & Zhong, 2014).
When the predictive relationships between social anxiety and alexithymia were examined in adolescents, it was determined that there was a positive linear relationship. In other words, findings reveal that as adolescents increase their social anxiety level, alexithymia will increase. Alexithymic features emerge in close connection with difficulties in accepting their own feelings and social anxiety features (Edel et al., 2010). In addition, alexithymia is a common personality trait in Turkish patients with social anxiety disorders (Solmaz, Sayar, Ozer, Ozturk, & Acar, 2000). There are studies supporting the findings obtained in this study. Kaur and Kaur (2015) was found that social anxiety was positively related to alexithymia in adolescents. Social anxiety symptoms are also significantly related to alexithymia in different samples (Evren & Evren, 2007;Panayiotou, Leonidou, Constantinou, & Michaelides, 2008). They found a positive relationship between anxiety and alexithymia jel.ccsenet.org Journal of Education and Learning Vol. 9, No. 2; 2020 (Berthoz, Consoli, Perez-Diaz, & Jouvent, 1999;Fukunishi, Kikuchi, Wogan, & Takubo, 1997;Karukivi et al., 2010;Nakao & Takeuchi, 2018;Stewart, Zvolensky, & Eifert, 2002). With an increase in anxiety complaints in alexithymic individuals, the "Difficulty in Defining Emotions" dimension increased (Motan & Gencoz, 2007). Again, the "Difficulty in Defining Emotions" subscale was found to be associated with social anxiety (Suslow, 1998). The sub-dimension of anxiety and difficulty in recognizing emotions was associated (Evren, Evren, Dalbudak, & Çakmak, 2008;Marchesi, Brusamonti, & Maggini, 2000). As a result, internet addiction and alexithymic features may increase in socially anxious adolescents.