Development of the Internet was prompted by the desire and need to store, share, and easily access information, and it has become an ever-growing global communication tool (Arısoy 2009; Douglas et al. 2008; Yalçın 2003). The rapid growth of technology since the 1990s has led to exponential increases in the number of Internet users worldwide. In 2012, 34.3% of the world’s population (2,405,518,376 people) and 36.4% of the population in Turkey (34,455,000 people) were accessing the Internet (internetworldstats 2015; ITU World Telecommunication 2014).

While the Internet serves a large population, the majority of its users are young people (Cömert and Kayıran 2010; Wang et al. 2013; Kelleci et al. 2009). According to 2015 data from TUİK (Turkey Statistical Institute), the 16–24 age group has the highest internet usage (84.3%) of any in Turkey. The Internet offers potential advantages to young people who are more open to learning and are more inclined to accept and implement innovations (Kabakci et al. 2008; Canbek and Sağıroğlu 2007; Johnson 2010). However, many studies have pointed out that the Internet is often used improperly and that computer and Internet use are associated with school absenteeism and decreases in school success (Kim et al. 2006; Valcke et al. 2007; Chang and Law 2008; Balta and Horzum 2008). Some research has shown that improper use of the Internet has harmful effects—especially on children and adolescents—and causes some physical, social, and psychological problems (Freeman 2008; Kim et al. 2006; Chang and Law 2008; Mittal et al. 2007; Xu et al. 2014). Children who spend too much time at their computers face challenges with physical, social, and psychological development (Canbek and Sağıroğlu 2007; Kurt et al. 2014). As a result, the Internet can be viewed as both beneficial to young people and the potential cause of irreparable damage to them (Kurt et al. 2014).

Some Internet users are identified as being “Internet addicted” or “problematic/pathological Internet users”. Although official diagnostic criteria do not yet exist, Internet addiction (IA) can be defined as the excessive, obsessive-compulsive, uncontrollable, tolerance-causing use of the Internet, which also creates significant distress and impairments in daily functioning (Young 1998, 1999). The prevalence of IA has been reported to vary between 1.5 and 8.2% worldwide (Petersen et al. 2009). Internet addiction has emerged as a burgeoning problem among children, adolescents, and young adults (e.g., college/university students) (Xu et al. 2014; Jie et al. 2014).

Internet addiction can be recognized by a person’s inability to overcome the desire for excessive use of the Internet, the loss of importance of the time spent not connected to the Internet, extreme irritability and aggression in deprivation, and the gradual deterioration of the person’s work, social, and family life (Davis 2001; Young 1999). According to Young (1999, 2007), IA leads to the significantly increased use of computers and the Internet, school failure, social and occupational disruptions and failures, an uncontrollable desire to use the Internet almost every night, and the feeling of fatigue which continues the next day. In 1995, Goldberg first used the term “Internet addiction” to describe a distinct psychiatric disorder. Today, studies have focused on establishing diagnostic criteria for Internet addiction, and it is under the section of compulsive-impulsive spectrum disorders in the Diagnostic and Statistical Manual of Mental Disorders-V (DSM-V). It is described as a common disorder in the DSM-V criteria (Köroğlu 2013; Block 2008).

Children and adolescents are more susceptible to the adverse effects of the Internet, especially in cases of excessive use, because their cognitive, emotional, and social development processes are not yet complete (Valcke et al. 2007). Children are starting to use the Internet at increasingly younger ages, causing concern in the scientific community (Yalçın 2003; Valcke et al. 2007; National Center for Educational Statistics 2007). Mood swings, a tendency toward depression, thought distortions, anger, and hostile feelings escalate during early adolescence, and they are expected to decline toward the end of adolescence. The effects of excessive Internet use may endure throughout this period because of these tendencies (Tahiroğlu et al. 2010).

Studies have examined the purpose of Internet use, activities on the Internet, Internet experiences, how long the Internet is used, and the personal characteristics of adolescents, along with their social and psychological well-being (Koç 2011; Mittal et al. 2007; Morahan-Martin and Schumacher 2003; Tam et al. 2007).

Many of these studies show that excessive Internet use is connected to psychosocial health. There is a positive relationship between problematic Internet use/IA and depression (Mittal et al. 2007; Tam et al. 2007), and research shows that levels of depression and suicidal ideation are higher in adolescents with IA (Kim et al. 2006). In Turkey, Özcan and Buzlu (2005) and Esen and Siyez (2011) reported a positive relationship between Internet addiction and the loneliness and depression levels of university students and adolescents, and a negative relationship between Internet addiction and the social support levels of these groups. Koç (2011) reported a positive relationship between the levels of Internet addiction and the psychiatric symptoms of young people. Literature reviews have revealed that the number of studies conducted on computer use by children and adolescents has increased in Turkey in recent years, but the number of studies on problematic Internet use/Internet addiction is limited (Balta and Horzum 2008; Canbek and Sağıroğlu 2007; Özcan and Buzlu 2005, 2007; Tahiroğlu et al. 2010).

It is necessary to determine the effects of internet addiction on the mental health of children and adolescents and to take necessary precautions in this regard. Adolescents spend a lot of time at school. To prevent Internet addiction among adolescents, it is important that health care professionals inform school administrators, teachers, and students’ parents about problematic behaviors.

For this reason, researchers carried out this descriptive study to determine problematic Internet use among high school students and the connections between that and the students’ family characteristics, social support, loneliness, and depression levels. The study addresses four main questions:

  • Do the family characteristics of young people affect their Internet use?

  • Is there a relationship between young people’s Internet use and their levels of loneliness?

  • Is there a relationship between young people’s Internet use and their levels of depression?

  • Is there a relationship between young people’s Internet use and their perceived levels of social support?

Methods

For this study, a descriptive design was used to determine problematic Internet use among high school students. The participants of this study were chosen from a pool of 8346 high school students in the 9th, 10th, 11th, and 12th grades in 18 secondary education institutions (affiliated with the Directorate of National Education) in the province of Edirne in the western part of Turkey during the 2011–2012 academic year. The study sample consisted of 881 students selected according to the 95% confidence interval, class, and gender, using simple random sampling methods. Sampling was done using a random number table with the students’ class lists.

Ethical Approach

Permission No. 04/01 (dated 29/12/2010) from the Scientific Research Evaluation Commission of the Faculty of Medicine at Trakya University and Permission No. 28665 (dated 17/02/2011) from the Edirne Provincial Directorate of National Education were obtained for the study. The data were obtained from students who volunteered to participate in the study. The purpose of the study was explained to the students by the researchers before the application; the students were asked not to write their names on the forms, and they were assured that the data collected would be kept confidential and used only for scientific purposes.

Data Collection

Procedures

The administrators of the 18 schools in which the study was to be conducted were interviewed prior to the implementation of the study and appointment dates were set up. Researchers went to the relevant schools on the appointed days. Classes were selected by the school management. Students selected by the simple random sampling method were gathered in the classes. The purpose of study was explained by researchers. Oral consent of the students was obtained. Any questions the students had were answered. The questionnaires were given to the students. The students had 20–25 min to complete the questionnaires. Data was collected in face-to-face interviews.

Data Collection Instruments

Researchers collected data using a questionnaire, the Online Cognition Scale (OCS), the Depression Scale for Children (CDS), the UCLA Loneliness Scale (ULS), and the Multidimensional Perceived Social Support Scale (MPSSS).

  • The Questionnaire. In accordance with the literature, researchers prepared a questionnaire that consisted of 28 questions regarding the sociodemographic characteristics of the adolescents and their families (age, gender, family type, family economic status, number of siblings, parents’ education, parents’ ages, parents’ occupations) and the Internet use of the adolescents (having their own computers, having Internet connections at home, the duration of Internet use, their intentions for using the Internet, the existence of rules in the family relating to Internet use).

  • Online Cognition Scale (OCS). Davis et al. (2002) developed the OCS to assess problematic Internet use and four sub-levels thereof: diminished impulse control, loneliness/depression, social comfort, and distraction. Özcan and Buzlu (2005) studied its validity and reliability in Turkey and an internal consistency coefficient of 0.91 was found. In this study, the internal consistency coefficient of the scale was 0.94. The OCS consists of 36 items and is a seven-point Likert-type scale that ranges from Strongly Disagree (one point) to Strongly Agree (seven points). Higher scores are considered to be indicative of problematic use.

  • UCLA Loneliness Scale (ULS). This scale was developed by Russell et al. (1978). Validity and reliability studies were conducted by Demir (1989) in Turkey and an internal consistency coefficient of 0.94 was found. In this study, the internal consistency coefficient of the scale was 0.77. The ULS consists of 20 items that assess, on a quartile Likert-type scale, how often feelings and thoughts related to social relations are experienced. The lowest and highest possible scores to be received from the scale are 20 and 80, respectively. Individuals with scores of 44 and above are defined as lonely.

  • Depression Scale for Children (CDS). Kovacs (1980) developed the CDS based on the Beck Depression Inventory, and Öy (1991) conducted its validity and reliability studies in Turkey. An internal consistency coefficient of 0.86 was found. In this study, the internal consistency coefficient of the scale was 0.84. The scale consists of 27 items. Each item comprises three sentences from which children can choose in evaluating their last 2 weeks. Answers are scored between zero and two. An overall depression score is obtained by collecting the individual scores tallied from the 27 items, making the highest score that can be received on the scale 54. A higher total score indicates a higher level or severity of depression. Children with scores greater than 19 are classified as depressed.

  • Multidimensional Perceived Social Support Scale (MPSSS). This 12-item scale was developed by Zimet et al. (1988). Eker and Arkar (1995) tested its validity and reliability in Turkey and an internal consistency coefficient of 0.86 was found among university students. In this study, the internal consistency coefficient of the scale was 0.88. It is a Likert-type scale of seven points ranging from Absolutely No to Absolutely Yes. The lowest and highest possible scores that can be received from the entire scale are 12 and 84, respectively. Higher scores indicate higher levels of perceived social support.

Data Analysis

The data were analyzed using the SPSS 20.00 (License No. 10240642) program by the Department of Biostatistics at the Trakya University Faculty of Medicine. The data relating students’ sociodemographic and internet usage were analyzed by percent, mean, and standard deviation. The normal distribution of the data was analyzed using the Kolmogorov-Smirnov test. In comparing the average scores of scale between the characteristics of the students and their families, t test and Mann Whitney U tests were used in the two groups, and one-way (ANOVA) and Kruskal-Wallis tests were used in the triple groups. Spearman’s correlation analysis was used for comparison of the students’ scores of scale. The results obtained were evaluated at a confidence interval of 95% and significance level of p < 0.05.

Results

This study was undertaken to examine the relationship between problematic Internet use among high school students and the students’ family characteristics, social support, loneliness, and depression levels.

Sociodemographic Characteristics of Students

Students in the study group ranged in age from 14 to 19 years, with the average being 16.52 ± 1.19. Of the 881 students, 51.2% were male, and 30.3% were in the 9th grade. One-third of the students’ mothers were primary school graduates (36.6%), and a majority were housewives (67.4%); a third of the students’ fathers were high school graduates (32.3%), and most were working (83.2%). A majority of the families were at the middle socioeconomic level (62.4%). Most of the students had siblings (84.4%), and slightly over half had one sibling only (57.5%) (Table 1).

Table 1 Sociodemographic characteristics of students (N = 881)
Full size table

Characteristics of Students Related to Internet Use

The students had been using the Internet for an average of 4.93 ± 2.35 years. The average duration of their daily Internet use was 2.29 ± 1.18 h, and the average duration of their weekly Internet use was 14.14 ± 12.03 h. Among the students, 78.8% had access to a computer at home, 60% had their own computers, 77.8% of the students used the Internet at home, and a majority did so in the evening (75.3%) (Table 2).

Table 2 Characteristics of students related to computer and internet use (N = 881)
Full size table

Scale Scores of Students

The average OCS score of the students was 92.24 ± 38.95. A positive correlation was found between the students’ OCS scores and their CDS (r = 0.231, p < 0.021) and ULS (r = 0.270, p < 0.001) scores, and a significant negative correlation was shown between the students’ OCS scores and their MPSSS scores (r = −0.151, p < 0.001) (Table 3).

Table 3 Relationship between the students’ OCS, CDS, ULS, and MPSSS scores
Full size table

A statistically significant relationship was found between the students’ genders and grade levels and their OCS mean scores (p < 0.001). The OCS mean score of the male students was higher than that of the female students, as the male students displayed more problematic Internet use. Further analysis established that the OCS mean score of the 12th grade students was lower than that of the 9th grade (p < 0.001), 10th grade (p = 0.002), and 11th grade students (p = 0.019). Thus, the 12th grade students had less problematic Internet use (Table 4).

Table 4 OCS mean scores of students by some characteristics (N = 881)
Full size table

A statistically significant relationship was found between the students who had their own computers and the OCS mean scores (p = 0.003). The OCS mean score of the students who had their own computers was higher, indicating greater problematic Internet use (Table 4).

A statistically significant relationship was found between what time of day the students used the Internet and the OCS mean score (p = 0.031). Further analysis showed that the OCS mean score of the students who used the Internet in the mornings was higher than that of the students who used the Internet during the daytime (p = 0.022). Similarly, the OCS average of the students who used the Internet in the evenings was higher than that of the students who used the Internet in the daytime (p = 0.018) (Table 4). The students who used the Internet in the mornings and evenings showed most problematic Internet use.

A statistically significant relationship was found between the students’ place of Internet access and the OCS mean score (p = 0.028). Upon further analysis, the OCS mean score of the students who used the Internet at home was greater than that of those who used the Internet at an Internet café (p = 0.047). The OCS mean score of the students who used the Internet at Internet cafés was lower than that of the students who used the Internet from their mobile phones (p = 0.027), and the OCS mean score of the students who used the Internet at school was also lower than that of the students who used the Internet on their mobile phones (p = 0.096). The students who used the Internet at home showed more problematic Internet use than those who used the Internet at Internet café. The students who used the Internet via their mobile phones showed more problematic Internet use than those who used the Internet at an Internet café and at school (Table 4).

There was a negative relationship between the students’ OCS scores and ages (r = −0.114, p = 0.001), and a positive relationship between their mother’s education level (r = 0.094; p = 0.005), their number of siblings (r = 0.074; p = 0.044), and their daily Internet use (r = 0.307; p < 0.001) (Table 5).

Table 5 Relationship between some characteristics and students’ OCS scores
Full size table

Discussion

This study examined the relationship between problematic Internet use among high school students and the students’ family characteristics, social support, loneliness, and depression levels. The average number of years students had been using the Internet was 4.93 ± 2.35. The mean of the daily duration of their Internet use was 2.29 ± 1.18 h (min: 1 h, max: 8 h), and 11.2% of the students used the Internet 4–5 h or more per day. Özcan and Buzlu (2007) reported that the average number of years the Turkish students they studied had been using the Internet was 3.17 ± 1.51 and Özcan and Buzlu (2005) stated that most of the students spent an average of 2.5 h on the Internet per week. Young and Rodgers (1998) reported that the average time problematic users spent on the Internet was 38.5 h a week while healthy users averaged 4.9 h a week. Jang and Ji (2012) found that online chatting and an increased duration of daily Internet use were associated with Internet addiction. The literature has shown that children are beginning to use the Internet at a younger age and that young people are at risk in this regard. The results of this study also support this finding.

The mean Online Cognition Scale score of the students was found to be 92.24 ± 38.95. Özcan and Buzlu’s (2007) study of students found a mean Online Cognition Scale score of 84.64 ± 33.50. The higher Online Cognition Scale scores in this study can be attributed to the increased use of the Internet in parallel with the development of Internet technology, increased access speeds, and the fact that the participants of this study are adolescent high school students. In addition, an important finding of this study is that the age of children using the Internet has decreased. In two Chinese studies, the prevalence of Internet addiction in adolescents was reported to be 7.5% (Wang et al. 2013) and 6% (Jie et al. 2014). In a study undertaken in the Netherlands, Kuss et al. (2013) declared that 3.7% of adolescents were Internet addicted, and that computer games and social networking sites increased the risk of Internet addiction.

In this study, male students had more problematic Internet (Table 4). Özcan and Buzlu (2007) stated that males spent more time on the Internet than females. Kelleci et al. (2009) observed that male high school students were more likely to use the Internet five or more hours per day, and Jang and Ji (2012) also reported that Internet addiction was more common among males. In studies conducted in China (Wang et al. 2013), in India (Goel et al. 2013), and in Iran (Ahmadi 2014), men were reported to have more problematic Internet use.

The current study discovered that problematic Internet use was lower among the 12th grade students (Table 4). In parallel with this finding, there was a negative correlation between the students’ Online Cognition Scale scores and their ages (Table 5). Studies indicate that the age at which children start using the Internet is declining (Yalçın 2003; Valcke et al. 2007; National Center for Educational Statistics 2007). The fact that the 12th grade students showed less problematic Internet use can be explained by the students’ ages. Similarly, Şahin (2011) and Yang and Tung (2004) found a negative relationship between class-age and Internet addiction levels. The increased duration of Internet use and addiction as the age decreases can be explained by the fact that children are being introduced to the Internet at younger ages. Şaşmaz et al. (2013) reported in their study conducted among high school students that 15.1% of the participants exhibited Internet addiction disorder and that the level of Internet addiction was higher in the first- and second-class students.

In this study, the problematic Internet use of high school students increased as the duration of their daily Internet use increased (Table 5). Andreou and Svoli (2013) reported that the extent of Internet usage was an important indicator for all aspects of Internet addiction. Several recent studies have noticed a positive correlation between Internet addiction levels and Internet use durations (Yang and Tung 2004; Şahin 2011; Wang et al. 2013).

The students’ loneliness and depression levels intensified, and their levels of perceived social support decreased as their Internet use increased (Table 5). Özcan and Buzlu (2007) reported that the loneliness and depression levels of university students rose, and their levels of perceived social support decreased as their OCS scores increased. Andreou and Svoli (2013) presented that depression was an important determinant of Internet addiction. Mittal et al. (2007) discovered that there was a positive correlation between Internet use and depression scores. Ybarra (2004) reported that young people with depressive symptoms used the Internet more intensely and communicated more online. Morahan-Martin and Schumacher (2003) found that lonely students used the Internet and e-mail more often; they had more online friends; they used the Internet to improve their moods and for emotional support; and the Internet caused impairments in their daily functioning. Jang et al. (2008) reported that obsessive-compulsive and depressive symptoms affected Internet addiction as independent factors, and Yen et al. (2014) discovered that adolescents with Internet addiction displayed more depression and social phobias. Koç (2011) reported that university students who used the Internet for 6 h a day had more psychiatric symptoms, and that there was a positive correlation between daily Internet use and the degrees of psychiatric symptoms. In the present study, as the levels of the students’ problematic Internet use increased, their levels of perceived social support decreased (Table 5). In other words, the adolescents with higher levels of perceived social support used the Internet less. One study implied that male adolescents having problems with their parents is a factor directly affecting Internet addiction (Jang and Ji 2012). In another study, it was reported that poor mother-adolescent relationships had an influence on Internet addiction, while poor father-adolescent relationships had a lesser influence on Internet addiction (Xu et al. 2014). The same study asserted that maternal factors played an important role in guiding adolescents in their Internet use. Ahmadi (2014) found that the Internet addiction levels of students with poor family relationships were higher. Jie et al. (2014) noted that Internet addiction was associated with students’ interpersonal and school-related problems, and Xu et al. (2014) reported that Internet addiction was more prevalent among lonely adolescents.

The mother-child relationship is important in the development of adolescent behaviors. Especially in Turkish culture, mothers assume more responsibility in raising the children, and fathers work more outside the home. Ahmadi (2014) reported that gender, age, the mother’s occupation, the family’s financial situation (both high and low), and loyalty to religious beliefs were associated with students’ Internet addictions.

The problematic Internet use of students who had their own computers was higher (Table 4). However, having an Internet connection at home did not affect problematic Internet use (Table 2). In this study, most students had their own computers and they used the ınternet at home and in the evenings. Wang et al. (2013) found that Internet addiction was higher among adolescents who had their own computers and a computer at home, and Goel et al. (2013) found that most of the addicts used the Internet in the evenings and at night. In the current study, the students who used the Internet in the mornings and evenings displayed higher problematic Internet use.

The levels of problematic Internet use among the students increased as the mothers’ education levels increased. Ahmadi (2014) discovered that Internet addiction was greater among male and female adolescents whose parents had doctorates. That same study found that Internet addiction was more prevalent among children whose mothers worked. Xu et al. (2014) reported that the socioeconomic status of a family had a moderate influence on the level of an adolescent’s Internet use but it did not affect Internet addiction, and the level of Internet use by adolescents with higher socioeconomic statuses was lower than that of adolescents with lower socioeconomic statuses.

Limitations

The study was limited to schools of Edirne, Turkey. Thus, study results are limited to Turkey (Edirne country) population only. The study could be conducted in large settings to validate and generalize its findings. A similar research can be taken with multi-setting approach.

Conclusion

It is important to recognize the factors affecting the levels of Internet use among high school students in order to solve the issues associated with problematic use. This study revealed that gender, grade level, time of Internet use, computer ownership, and the Internet use location affected the students’ mean OCS scores. Male students, younger students, those who used the Internet in the mornings and the evenings, and those who accessed the Internet from a computer at home or from their own computer or mobile phone displayed more problematic Internet use. This research also showed that as the students’ maternal education level, number of siblings, and duration of Internet use increased, their problematic Internet use increased. Additionally, as the students’ problematic Internet use increased, their depression and loneliness scores increased, and their levels of perceived social support decreased.

It would be worthwhile to organize internet usage education program in cooperation with relevant experts. These should be informed by the study results and take the factors influencing students’ problematic Internet use into account. Sharing the results of the study with school administrators, students, and parents would also raise awareness in this regard, so that problematic Internet use and the factors affecting it will be better understood.