Journal of Adolescent Health 56 (2015) 300e306
www.jahonline.org Original article
Stress Moderates the Relationship Between Problematic Internet Use by Parents and Problematic Internet Use by Adolescents Lawrence T. Lam, M.P.H., Ph.D. a, b, *, and Emmy M. Y. Wong, Ph.D. b a b
Discipline of Paediatrics and Child Health, The Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia Department of Health and Physical Education, The Hong Kong Institute of Education, Hong Kong SAR, China
Article history: Received May 23, 2014; Accepted October 16, 2014 Keywords: Problematic Internet use; Internet addiction; Stress; Parent; Adolescents; Dyad study
A B S T R A C T
Purpose: Based on the theoretical framework of Problem Behavior and Stress Reduction theories for problematic Internet use (PIU), this study aimed to investigate the relationship between parental PIU and the PIU among adolescents taking into consideration the stress levels of young people. Methods: This was a population-based parent and adolescent dyad health survey utilizing a random sampling technique. PIU for both parents and adolescents was measured by the Internet addiction test designed by Young. The stress level of adolescents was assessed using the stress subscale of the Depression Anxiety Stress Scale (DASS). Data were analyzed using logistic regression modeling techniques with adjustment for potential confounding factors with analysis on the modification effect of stress levels on the relationship between parent and adolescent PIU. Results: Of the total 1,098 parent and adolescent dyads with usable information, 263 adolescents (24.0%) and 62 parents (5.7%) could be classified as moderate and severe problematic users of the Internet. About 14% (n ¼ 157) of adolescents could be classified with moderate-to-severe stress. Regression analysis results suggested a significant interaction between parental PIU and adolescents’ stress levels on adolescent PIU. Stratified regression analyses by stress level resulted in a significant parent and adolescent PIU relationship in the low stress group (odds ratio, 3.18; 95% confidence interval 1.65e6.14). However, the association between parent and adolescent PIU in the high stress group became insignificant. Conclusions: There was a significant parent and adolescent PIU relationship; however, this relationship is differentially affected by the stress status of the adolescent. The direct implication of the results is that parental Internet use should also be assessed and included as part of the treatment regime for adolescents. Ó 2015 Society for Adolescent Health and Medicine. All rights reserved.
Problematic Internet use (PIU) is an emerging behavioral problem of concern related to human interactions with information and communication technologies, particularly among adolescents [1]. Although Internet gaming addiction has been included as an emerging disorder worthy for further * Address correspondence to: Lawrence T. Lam, M.P.H., Ph.D., The Hong Kong Institute of Education, 10 Lo Ping Road, Tai Po, N.T., Hong Kong SAR, China. E-mail address: [email protected] (L.T. Lam). 1054-139X/Ó 2015 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/j.jadohealth.2014.10.263
IMPLICATIONS AND CONTRIBUTION
Problematic Internet use by parents is related to problematic Internet use by their adolescent, and this relationship is moderated by the child’s stress level. Parental Internet use should be assessed when treating adolescents for these conditions.
consideration in the latest version of the DSM-5 [2], “Internet addiction” is yet to be fully recognized as an established disorder with some controversial debate still ongoing [3,4]. Different conceptual frameworks have been proposed as the theoretical basis for the understanding of PIU, particularly among adolescents [5,6]. Among these theoretical models, stress or anxiety reduction has been proposed as a possible explanatory theory for PIU [7]. According to this theory, the motivation for the
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
behavioral maintenance of PIU in “over users” is that the Internet is used as a means for stress or tension reduction [7]. Another theory, which has also received much attention, especially for PIU among adolescents, is Problem Behavior Theory [8e10]. The Problem Behavior Theory advocates that there are three main systemsdnamely personality, environment, and behavioraldin the conceptual structure of any problematic behaviors in young people [11,12]. The propensity for any involvement in problematic behaviors is determined by the balance between risk and protective factors among the three systems [12]. Moreover, the behavior system is also influenced by interactions between the personality and environment systems [12]. According to Jessor et al. [12] familial and parental factors, such as parental behaviors and attitudes, are the main focus of the environment system, and risk factors in the personality system include lack of achievement, alienation, and self-dissatisfaction. In terms of the relationship between stress and PIU among adolescents, a number of studies in different countries have been reported in the literature [13e16]. In a study among Chinese adolescents, it was reported that young people who had experienced recent stressful life events were 10 times as likely to be involved in PIU (odds ratio [OR], 10.0; 95% confidence interval [CI], 6.5e12.2) in comparison with those who had little stressful experiences [15]. A study among young people in Germany also reported that maladaptive coping behavior and stress were related to excessive computer use and gaming behaviors [16]. While applying the Problem Behavior Theory to the understanding of PIU among adolescents, particularly in the identification of important risk factors, the literature has yielded some valuable insight. In terms of personal factors, studies have demonstrated that PIU is associated with low academic achievement, dissatisfaction with the family, and other risky behaviors [8,9,15,17e19]. For familial and parental factors of PIU among adolescents, there is a small but growing volume of literature in this particular area of PIU research [20e26]. Among these studies, some common parental elements were identified as associated with PIU of the adolescent. These included dysfunctional family relationships, particularly the parent and adolescent relationship, and parental attitudes [20,21,24e26]. Although most of these were studies with large samples recruited using sound methodological and sampling techniques, all suffered from the same drawback in which parental information was collected through reporting from the adolescent and not from the parents per se. Moreover, none of these studies attempted to investigate the behavioral aspects of parents and the possible impact on adolescent PIU. As suggested by Problem Behavior Theory, parental behavior is an important factor that may exert a significant influence on the adolescent’s behavior. It would be prudent to postulate that there is a relationship between parent and adolescent PIU. As such, there is a need to further explore the relationship between parental Internet behavior, particularly parental PIU, and PIU of their offspring. Although drawing on the conceptual thinking of Problem Behavior and Stress Reduction theories, this study aims to bridge the knowledge gap through examining the relationship between parental PIU and PIU among adolescents. Furthermore, it also aims to investigate the role of adolescent stress in the relationship between parent and adolescent PIU, with parental information obtained directly from parents. Based on the aforementioned theories, it is hypothesized that there is a significant association between parental PIU and the PIU of their
301
offspring. It is further hypothesized that the stress status of the adolescent may also play an important role in the adolescent’s PIU such that it modifies the parent and adolescent PIU relationship. Another possible hypothesis is that the relationship between parental PIU and the PIU of their offspring is mediated through the stress level of adolescents such that parental PIU is in some ways associated with stress in adolescents and in turn influences the Internet behavior of the offspring. Although this may be a theoretically possible hypothesis, it has not been suggested in the literature with indicative empirical evidence. Hence, this hypothesis will not be entertained and tested in this study. Methods This parent and adolescent dyad cross-sectional survey study was conducted in Hong Kong in March 2014 among 13- to 17-year-old high school adolescents. Institute ethics approval for the study was granted by the Hong Kong Institute of Education. The sample was generated from the total adolescent population of adolescents who attended high schools within a specific local school region. These high schools were registered with the Hong Kong Education Bureau, the governmental body responsible for the education of all pretertiary adolescents in Hong Kong. Two schools were randomly selected from the list of registered high schools to be the target schools. To formulate the final sample, a class was also randomly selected from each grade, from Grade 7 to 11, with all adolescents and their parents in the class invited to participate in the study. The study was conducted on campus at the two schools via a health survey carried out within a period of a few weeks. Selected adolescents and their parents were invited to participate in the study with endorsements from the school. Information on the study was provided to selected adolescents and their parents via school principals and their teachers. Informed consent was obtained from all participating adolescents and parents with a signed consent form indicating willful participation of the parent and adolescent dyad. After signing the consent form, both parents and adolescents were invited to fill in the Parent’s and Adolescent’s Health Survey Questionnaires, respectively. Completed questionnaires were returned to the school in a sealed envelope to safeguard the confidentiality of respondents. Measures The Parent’s and Adolescent’s Health Survey Questionnaires included similar questions with some specifically designed for parents or adolescents. PIU was assessed by the Internet addiction test (IAT) also known as the Young’s Internet Addiction Scale designed by Young [27]. The IAT is a 20-item self-reported scale, and the design was based on the concepts and behaviors exhibited by pathological gamblers as defined by the DSM-4 diagnostic criteria. It includes questions that reflect typical behaviors of addiction. An example question is “How often do you feel depressed, moody, or nervous when you are off-line, which goes away once you are back on-line?” Respondents were asked to indicate their responses on a Likert scale ranging from 1 (rarely) to 5 (always). A study on the psychometric properties of the IAT suggested good reliability with Cronbach’s alpha values ranged from .82 to .54 for various factors [28]. Total scores were calculated with possible scores ranging from a minimum of 20 to a
302
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
maximum of 100. The severity of addiction was then classified according to the suggested cutoff scores with 20e49 points as “normal,” 50e79 points as “moderate,” and 80e100 points as “severe” [27]. For ease of analysis, the variable was dichotomized into two categories: “severe/moderate” and “mild/normal” for both parent and adolescent. The stress level of adolescents was measured using the stress scale from the Depression Anxiety Stress Scale (DASS) [29]. The DASS is a fully validated and commonly used instrument designed for the assessment of stress, depressive symptoms, and anxiety with good psychometric properties including strong reliability and validity. It has also been recommended for use among children and adolescents [30]. As suggested by the authors of the scale, the DASS was designed as a quantitative measure of distress along three axes; however, it was not meant to be a categorical assessment for clinical diagnosis [29]. Nevertheless, the scale could be useful for identifying individuals who were at high risk of mental health problems. In this study, the stress score of the individual was classified according to the recommended categorization, with 0e7 as “normal,” 8e9 as “mild,” 10e12 as “moderate,” and 13 or above as “severe or extremely severe.” For ease of analysis, the variable was also dichotomized into two categories: “mild/normal” or “moderate or above.” In accordance with Jessor’s Problem Behavior Theory, selfperception on life situation of adolescents is considered as an important variable in determining behavioral involvement. Hence, additional information was also gathered on the level of life satisfaction using the Multidimensional Adolescents Life Satisfaction Scale [31,32]. The Multidimensional Adolescents Life Satisfaction Scale was designed as a multidimensional profile of adolescents’ perception of their life satisfaction in areas that are important to them. These included specific domains, such as school, family, and friends. Respondents were asked to rate on a five-point Likert scale their agreement to each statement. Studies reported high internal consistency of all subscales ranging from .70 seconds to .90 seconds [33]. These scales also had reasonably high testeretest reliability within 2- and 4-week intervals with the intraclass correlation coefficient between .70 and .93 [33]. Other information collected in the adolescents’ survey included demographics, location of family residence, whether the adolescent was born in Hong Kong, and some details on the means and patterns of accessing the Internet. For parents, questions on sex, age, occupation, and access to the Internet were also included in the questionnaire. Data analysis Data were analyzed using the Stata v10.0 statistical software program (StataCorp., College Station, TX). Descriptive analyses were conducted using percentages, means, and standard deviation. Bivariate analyses were conducted to examine unadjusted relationships between the outcome variable (i.e., adolescent PIU) and variables of interest such as demographics and Internet usage of adolescents, adolescent stress levels, and parental PIU. With the main focus of analyses on the relationship between parent and adolescent PIU, potential confounding variables identified from the bivariate analyses were included in further logistic regression analyses. A significant bivariate result with a p < .10 was used as the selection criteria of potential confounding variables. The adjusted relationship between parent and adolescent PIU was further examined using the multiple logistic regression approach
with the calculation of the 95% CI. To investigate any possible effect modification between parental PIU and other variables on adolescent’s PIU, in particular adolescents’ stress levels, the interaction terms were tested using a Type I error rate of 1%. Should an interaction term between the exposure variable (parental PIU) and adolescent stress level be found significant in the regression model, further analyses on the relationship between parental PIU and adolescent PIU was conducted with stratification by stress levels. Different regression models were fitted to the data for the high and the low stress groups separately. Results A total of 1,098 parent and adolescent dyads were recruited and responded to the survey providing usable information and allowed matching of parent and adolescent data. This represented a response rate of 95.3% of parents completing the questionnaire with 54 adolescents without a response from parents. Comparisons between those adolescents with a respondent parent and those whose parents did not respond indicated no statistically significant differences in all demographics, including age, sex, grade, and place of birth. The parent and adolescent characteristics and outcome measures of the respondents were summarized in Table 1. The sample consisted of adolescents aged between 13 and 17 years old with slightly more than half aged 15 years or older (n ¼ 620, 56.5%) and 56% females (n ¼ 614). In terms of demographics, the majority of the adolescents were born locally (n ¼ 1,034, 94.3%) and currently lived in Hong Kong (n ¼ 1,056, 96.3%). In terms of means and patterns of Internet access, slightly more than half owned a room him/herself with Internet access (n ¼ 576, 52.5%), 694 (63.2%) owned a personal computer with Internet access, and 89.4% (n ¼ 982) reported to have a smartphone with Internet access as well. The majority of adolescents accessed the Internet everyday (n ¼ 764, 69.8%), and nearly 12% (n ¼ 129) reported accessing the Internet 4e6 days in the week before the survey. Slightly less than half (n ¼ 508, 46.3%) accessed the Internet using a computer, and 45.7% (n ¼ 501) used a mobile device. For the duration of Internet access, more than one-third (n ¼ 403, 37.9%) reported that on average, they spent 3 hours or more per day accessing the Internet, of which 197 (49.0%) accessed the Internet for playing online games. Seventy adolescents (6.4%) admitted to having visited a pornographic site for more than 5 minutes in the week before the survey, and slightly less than 100 (n ¼ 96, 8.8%) revealed that they had either sent or received offensive texts or images online in the week before the survey. In terms of psychosocial factors, about 14% (n ¼ 157) of respondents could be classified with moderate-to-severe stress and the mean life satisfaction score of the sample was 68.4 (SD ¼ 9.65). For parents, the majority were females (n ¼ 814, 74.5%) with more than half aged 45 years and older (n ¼ 602, 56.5%). In contrast to their offspring, more than half (n ¼ 641, 58.5%) of parents were born outside Hong Kong and slightly more than a quarter (n ¼ 317, 29%) of parents were employed as professionals, semiprofessionals, or nonprofessionals. In terms of the PIU, 263 adolescents (24.0%) could be classified as moderate and severe, whereas 62 (5.7%) parents could be considered as having a moderate-to-severe problem. The bivariate relationships between PIU of adolescents, other variables of interest, and parental PIU were examined. The results were also summarized in Table 1. As shown, adolescent PIU was significantly associated with parental PIU unadjusted for
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
303
Table 1 Frequency (%) or mean (standard deviation) of the characteristics of adolescent and parent dyad by adolescent’s problematic Internet use (PIU) status and results of associations (N ¼ 1,098) Characteristics of respondent
Adolescent characteristics Sex Male Female Age group 15 years or older
www.jahonline.org Original article
Stress Moderates the Relationship Between Problematic Internet Use by Parents and Problematic Internet Use by Adolescents Lawrence T. Lam, M.P.H., Ph.D. a, b, *, and Emmy M. Y. Wong, Ph.D. b a b
Discipline of Paediatrics and Child Health, The Sydney Medical School, The University of Sydney, Sydney, New South Wales, Australia Department of Health and Physical Education, The Hong Kong Institute of Education, Hong Kong SAR, China
Article history: Received May 23, 2014; Accepted October 16, 2014 Keywords: Problematic Internet use; Internet addiction; Stress; Parent; Adolescents; Dyad study
A B S T R A C T
Purpose: Based on the theoretical framework of Problem Behavior and Stress Reduction theories for problematic Internet use (PIU), this study aimed to investigate the relationship between parental PIU and the PIU among adolescents taking into consideration the stress levels of young people. Methods: This was a population-based parent and adolescent dyad health survey utilizing a random sampling technique. PIU for both parents and adolescents was measured by the Internet addiction test designed by Young. The stress level of adolescents was assessed using the stress subscale of the Depression Anxiety Stress Scale (DASS). Data were analyzed using logistic regression modeling techniques with adjustment for potential confounding factors with analysis on the modification effect of stress levels on the relationship between parent and adolescent PIU. Results: Of the total 1,098 parent and adolescent dyads with usable information, 263 adolescents (24.0%) and 62 parents (5.7%) could be classified as moderate and severe problematic users of the Internet. About 14% (n ¼ 157) of adolescents could be classified with moderate-to-severe stress. Regression analysis results suggested a significant interaction between parental PIU and adolescents’ stress levels on adolescent PIU. Stratified regression analyses by stress level resulted in a significant parent and adolescent PIU relationship in the low stress group (odds ratio, 3.18; 95% confidence interval 1.65e6.14). However, the association between parent and adolescent PIU in the high stress group became insignificant. Conclusions: There was a significant parent and adolescent PIU relationship; however, this relationship is differentially affected by the stress status of the adolescent. The direct implication of the results is that parental Internet use should also be assessed and included as part of the treatment regime for adolescents. Ó 2015 Society for Adolescent Health and Medicine. All rights reserved.
Problematic Internet use (PIU) is an emerging behavioral problem of concern related to human interactions with information and communication technologies, particularly among adolescents [1]. Although Internet gaming addiction has been included as an emerging disorder worthy for further * Address correspondence to: Lawrence T. Lam, M.P.H., Ph.D., The Hong Kong Institute of Education, 10 Lo Ping Road, Tai Po, N.T., Hong Kong SAR, China. E-mail address: [email protected] (L.T. Lam). 1054-139X/Ó 2015 Society for Adolescent Health and Medicine. All rights reserved. http://dx.doi.org/10.1016/j.jadohealth.2014.10.263
IMPLICATIONS AND CONTRIBUTION
Problematic Internet use by parents is related to problematic Internet use by their adolescent, and this relationship is moderated by the child’s stress level. Parental Internet use should be assessed when treating adolescents for these conditions.
consideration in the latest version of the DSM-5 [2], “Internet addiction” is yet to be fully recognized as an established disorder with some controversial debate still ongoing [3,4]. Different conceptual frameworks have been proposed as the theoretical basis for the understanding of PIU, particularly among adolescents [5,6]. Among these theoretical models, stress or anxiety reduction has been proposed as a possible explanatory theory for PIU [7]. According to this theory, the motivation for the
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
behavioral maintenance of PIU in “over users” is that the Internet is used as a means for stress or tension reduction [7]. Another theory, which has also received much attention, especially for PIU among adolescents, is Problem Behavior Theory [8e10]. The Problem Behavior Theory advocates that there are three main systemsdnamely personality, environment, and behavioraldin the conceptual structure of any problematic behaviors in young people [11,12]. The propensity for any involvement in problematic behaviors is determined by the balance between risk and protective factors among the three systems [12]. Moreover, the behavior system is also influenced by interactions between the personality and environment systems [12]. According to Jessor et al. [12] familial and parental factors, such as parental behaviors and attitudes, are the main focus of the environment system, and risk factors in the personality system include lack of achievement, alienation, and self-dissatisfaction. In terms of the relationship between stress and PIU among adolescents, a number of studies in different countries have been reported in the literature [13e16]. In a study among Chinese adolescents, it was reported that young people who had experienced recent stressful life events were 10 times as likely to be involved in PIU (odds ratio [OR], 10.0; 95% confidence interval [CI], 6.5e12.2) in comparison with those who had little stressful experiences [15]. A study among young people in Germany also reported that maladaptive coping behavior and stress were related to excessive computer use and gaming behaviors [16]. While applying the Problem Behavior Theory to the understanding of PIU among adolescents, particularly in the identification of important risk factors, the literature has yielded some valuable insight. In terms of personal factors, studies have demonstrated that PIU is associated with low academic achievement, dissatisfaction with the family, and other risky behaviors [8,9,15,17e19]. For familial and parental factors of PIU among adolescents, there is a small but growing volume of literature in this particular area of PIU research [20e26]. Among these studies, some common parental elements were identified as associated with PIU of the adolescent. These included dysfunctional family relationships, particularly the parent and adolescent relationship, and parental attitudes [20,21,24e26]. Although most of these were studies with large samples recruited using sound methodological and sampling techniques, all suffered from the same drawback in which parental information was collected through reporting from the adolescent and not from the parents per se. Moreover, none of these studies attempted to investigate the behavioral aspects of parents and the possible impact on adolescent PIU. As suggested by Problem Behavior Theory, parental behavior is an important factor that may exert a significant influence on the adolescent’s behavior. It would be prudent to postulate that there is a relationship between parent and adolescent PIU. As such, there is a need to further explore the relationship between parental Internet behavior, particularly parental PIU, and PIU of their offspring. Although drawing on the conceptual thinking of Problem Behavior and Stress Reduction theories, this study aims to bridge the knowledge gap through examining the relationship between parental PIU and PIU among adolescents. Furthermore, it also aims to investigate the role of adolescent stress in the relationship between parent and adolescent PIU, with parental information obtained directly from parents. Based on the aforementioned theories, it is hypothesized that there is a significant association between parental PIU and the PIU of their
301
offspring. It is further hypothesized that the stress status of the adolescent may also play an important role in the adolescent’s PIU such that it modifies the parent and adolescent PIU relationship. Another possible hypothesis is that the relationship between parental PIU and the PIU of their offspring is mediated through the stress level of adolescents such that parental PIU is in some ways associated with stress in adolescents and in turn influences the Internet behavior of the offspring. Although this may be a theoretically possible hypothesis, it has not been suggested in the literature with indicative empirical evidence. Hence, this hypothesis will not be entertained and tested in this study. Methods This parent and adolescent dyad cross-sectional survey study was conducted in Hong Kong in March 2014 among 13- to 17-year-old high school adolescents. Institute ethics approval for the study was granted by the Hong Kong Institute of Education. The sample was generated from the total adolescent population of adolescents who attended high schools within a specific local school region. These high schools were registered with the Hong Kong Education Bureau, the governmental body responsible for the education of all pretertiary adolescents in Hong Kong. Two schools were randomly selected from the list of registered high schools to be the target schools. To formulate the final sample, a class was also randomly selected from each grade, from Grade 7 to 11, with all adolescents and their parents in the class invited to participate in the study. The study was conducted on campus at the two schools via a health survey carried out within a period of a few weeks. Selected adolescents and their parents were invited to participate in the study with endorsements from the school. Information on the study was provided to selected adolescents and their parents via school principals and their teachers. Informed consent was obtained from all participating adolescents and parents with a signed consent form indicating willful participation of the parent and adolescent dyad. After signing the consent form, both parents and adolescents were invited to fill in the Parent’s and Adolescent’s Health Survey Questionnaires, respectively. Completed questionnaires were returned to the school in a sealed envelope to safeguard the confidentiality of respondents. Measures The Parent’s and Adolescent’s Health Survey Questionnaires included similar questions with some specifically designed for parents or adolescents. PIU was assessed by the Internet addiction test (IAT) also known as the Young’s Internet Addiction Scale designed by Young [27]. The IAT is a 20-item self-reported scale, and the design was based on the concepts and behaviors exhibited by pathological gamblers as defined by the DSM-4 diagnostic criteria. It includes questions that reflect typical behaviors of addiction. An example question is “How often do you feel depressed, moody, or nervous when you are off-line, which goes away once you are back on-line?” Respondents were asked to indicate their responses on a Likert scale ranging from 1 (rarely) to 5 (always). A study on the psychometric properties of the IAT suggested good reliability with Cronbach’s alpha values ranged from .82 to .54 for various factors [28]. Total scores were calculated with possible scores ranging from a minimum of 20 to a
302
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
maximum of 100. The severity of addiction was then classified according to the suggested cutoff scores with 20e49 points as “normal,” 50e79 points as “moderate,” and 80e100 points as “severe” [27]. For ease of analysis, the variable was dichotomized into two categories: “severe/moderate” and “mild/normal” for both parent and adolescent. The stress level of adolescents was measured using the stress scale from the Depression Anxiety Stress Scale (DASS) [29]. The DASS is a fully validated and commonly used instrument designed for the assessment of stress, depressive symptoms, and anxiety with good psychometric properties including strong reliability and validity. It has also been recommended for use among children and adolescents [30]. As suggested by the authors of the scale, the DASS was designed as a quantitative measure of distress along three axes; however, it was not meant to be a categorical assessment for clinical diagnosis [29]. Nevertheless, the scale could be useful for identifying individuals who were at high risk of mental health problems. In this study, the stress score of the individual was classified according to the recommended categorization, with 0e7 as “normal,” 8e9 as “mild,” 10e12 as “moderate,” and 13 or above as “severe or extremely severe.” For ease of analysis, the variable was also dichotomized into two categories: “mild/normal” or “moderate or above.” In accordance with Jessor’s Problem Behavior Theory, selfperception on life situation of adolescents is considered as an important variable in determining behavioral involvement. Hence, additional information was also gathered on the level of life satisfaction using the Multidimensional Adolescents Life Satisfaction Scale [31,32]. The Multidimensional Adolescents Life Satisfaction Scale was designed as a multidimensional profile of adolescents’ perception of their life satisfaction in areas that are important to them. These included specific domains, such as school, family, and friends. Respondents were asked to rate on a five-point Likert scale their agreement to each statement. Studies reported high internal consistency of all subscales ranging from .70 seconds to .90 seconds [33]. These scales also had reasonably high testeretest reliability within 2- and 4-week intervals with the intraclass correlation coefficient between .70 and .93 [33]. Other information collected in the adolescents’ survey included demographics, location of family residence, whether the adolescent was born in Hong Kong, and some details on the means and patterns of accessing the Internet. For parents, questions on sex, age, occupation, and access to the Internet were also included in the questionnaire. Data analysis Data were analyzed using the Stata v10.0 statistical software program (StataCorp., College Station, TX). Descriptive analyses were conducted using percentages, means, and standard deviation. Bivariate analyses were conducted to examine unadjusted relationships between the outcome variable (i.e., adolescent PIU) and variables of interest such as demographics and Internet usage of adolescents, adolescent stress levels, and parental PIU. With the main focus of analyses on the relationship between parent and adolescent PIU, potential confounding variables identified from the bivariate analyses were included in further logistic regression analyses. A significant bivariate result with a p < .10 was used as the selection criteria of potential confounding variables. The adjusted relationship between parent and adolescent PIU was further examined using the multiple logistic regression approach
with the calculation of the 95% CI. To investigate any possible effect modification between parental PIU and other variables on adolescent’s PIU, in particular adolescents’ stress levels, the interaction terms were tested using a Type I error rate of 1%. Should an interaction term between the exposure variable (parental PIU) and adolescent stress level be found significant in the regression model, further analyses on the relationship between parental PIU and adolescent PIU was conducted with stratification by stress levels. Different regression models were fitted to the data for the high and the low stress groups separately. Results A total of 1,098 parent and adolescent dyads were recruited and responded to the survey providing usable information and allowed matching of parent and adolescent data. This represented a response rate of 95.3% of parents completing the questionnaire with 54 adolescents without a response from parents. Comparisons between those adolescents with a respondent parent and those whose parents did not respond indicated no statistically significant differences in all demographics, including age, sex, grade, and place of birth. The parent and adolescent characteristics and outcome measures of the respondents were summarized in Table 1. The sample consisted of adolescents aged between 13 and 17 years old with slightly more than half aged 15 years or older (n ¼ 620, 56.5%) and 56% females (n ¼ 614). In terms of demographics, the majority of the adolescents were born locally (n ¼ 1,034, 94.3%) and currently lived in Hong Kong (n ¼ 1,056, 96.3%). In terms of means and patterns of Internet access, slightly more than half owned a room him/herself with Internet access (n ¼ 576, 52.5%), 694 (63.2%) owned a personal computer with Internet access, and 89.4% (n ¼ 982) reported to have a smartphone with Internet access as well. The majority of adolescents accessed the Internet everyday (n ¼ 764, 69.8%), and nearly 12% (n ¼ 129) reported accessing the Internet 4e6 days in the week before the survey. Slightly less than half (n ¼ 508, 46.3%) accessed the Internet using a computer, and 45.7% (n ¼ 501) used a mobile device. For the duration of Internet access, more than one-third (n ¼ 403, 37.9%) reported that on average, they spent 3 hours or more per day accessing the Internet, of which 197 (49.0%) accessed the Internet for playing online games. Seventy adolescents (6.4%) admitted to having visited a pornographic site for more than 5 minutes in the week before the survey, and slightly less than 100 (n ¼ 96, 8.8%) revealed that they had either sent or received offensive texts or images online in the week before the survey. In terms of psychosocial factors, about 14% (n ¼ 157) of respondents could be classified with moderate-to-severe stress and the mean life satisfaction score of the sample was 68.4 (SD ¼ 9.65). For parents, the majority were females (n ¼ 814, 74.5%) with more than half aged 45 years and older (n ¼ 602, 56.5%). In contrast to their offspring, more than half (n ¼ 641, 58.5%) of parents were born outside Hong Kong and slightly more than a quarter (n ¼ 317, 29%) of parents were employed as professionals, semiprofessionals, or nonprofessionals. In terms of the PIU, 263 adolescents (24.0%) could be classified as moderate and severe, whereas 62 (5.7%) parents could be considered as having a moderate-to-severe problem. The bivariate relationships between PIU of adolescents, other variables of interest, and parental PIU were examined. The results were also summarized in Table 1. As shown, adolescent PIU was significantly associated with parental PIU unadjusted for
L.T. Lam and E.M.Y. Wong / Journal of Adolescent Health 56 (2015) 300e306
303
Table 1 Frequency (%) or mean (standard deviation) of the characteristics of adolescent and parent dyad by adolescent’s problematic Internet use (PIU) status and results of associations (N ¼ 1,098) Characteristics of respondent
Adolescent characteristics Sex Male Female Age group 15 years or older
