This study is a multi-phase cross-sectional study that included individual interviews, focus group discussions, survey questionnaire, and biological analysis. It was a part of larger study that examined the role of autonomic nervous system in development of IGA among adolescent males [28, 33, 34]. The current study focuses on identifying categories of gaming reasons and their relationships to IGA risk and to biological indicators (NE and cortisol levels). Although IGA risk and plasma NE and cortisol levels are measured in our previous studies, the research question and study sample are different. For example, the current study included only adolescents currently engaged in internet gaming.
Participants and procedures
All study participants were male high school students (adolescents) who came to a regional health center in a city in South Korea in response to an advertisement about the study. Convenience and snowball sampling methods were used to recruit the sample. This study had three phases. In the first phase 15 adolescent males were interviewed individually; the second phase involved eight participants in a focus group discussion; and in the third phase 225 participants completed a questionnaire and provided a sample of blood.
Gaming reasons were generated in the first and the second phases of the study. Participants in the first phase were interviewed individually to describe gaming reasons. Participants in the second phase (focus group) were asked to affirm or revise/add to the gaming reasons obtained in the first phase. Both interviews and the focus group were conducted in a private room, and responses were incorporated into the questionnaire used in phase three. Participants in the third phase first completed the questionnaire in a private room and then blood samples were drawn. All participants fasted for 12 hours before blood sampling. In addition, participants were instructed not to smoke, drink caffeinated beverages, or engage in internet gaming for 24 hours prior to data collection. This study was approved by the Institutional Review Board of a University. Informed consent was obtained from all participants and their legal guardians.
The sample size for the third phase was determined using one-way ANOVA analysis based on a medium effect size 0.25 [12, 14], an alpha level of 0.05, and a power of 0.80 using the G-power software . A minimum sample size of 180 was estimated and it was determined that and our sample size of 225 should provide ample power to detect statistically significant findings. The study sample was limited to male participants because male adolescents are known to be more commonly addicted to internet gaming than are their female counterparts  and because reasons for gaming may differ by gender . The flowchart for sampling procedures is depicted in Figure 1.
Measures for phase three of the study consisted of demographic characteristics; internet gaming-related information, including gaming reasons; an IGA risk assessment scale; and biological indicators. These measures were assembled into a single questionnaire. The demographic and internet gaming-related items were generated from the literature (excluding gaming reasons) and content validity and reliability of the items were established by four content experts.
Demographic items included participant age as well as information about smoking, drinking, and sleep time. Data related to current smoking and to alcohol drinking were obtained using yes/no responses. Sleep time was obtained using two categories; six hours or more a day and less than six hours a day.
Internet gaming-related information
Internet gaming-related information included amount of time spent on internet gaming (minutes per day) and duration of internet gaming participation (years). Participants were asked to select one of four reasons that best described why they engaged in internet gaming. The categories included (1) entertainment, (2) getting along with friends (or friendship), (3) stress relief, and (4) habitual gaming. The four categories were derived from individual interviews with 15 participants in the first phase and from the focus group interview with 8 participants in the second phase of this study. These interviews were performed by the principal investigator to identify perceived internet gaming reasons expressed in the adolescents’ own words. In the first phase, participants were asked to describe as many gaming reasons as possible in response to the question, “Why do you play internet games?” The duration of each interview ranged from 5 to 30 minutes. Overall, 36 statements were derived and analyzed by two independent coders of the research team, they having experience with content analysis. In the second phase, the team verified statements derived in the first phase and conducted a 45-minute focus group interview in which 8 adolescents were asked about their agreement with the 36 statements and whether they could think of additional reasons for playing internet games. During the discussion, participants were instructed to respond to questions by providing the first thought or feeling coming to their mind based on their real experiences and views. Participants’ responses were simple and brief, e.g., “because it’s fun,” “to play with friends,” or “just to do.” The 32 statements generated from the focus group discussion were similar to those generated in the first phase. A total of 68 statements were categorized by the two independent coders of the research team initially and then validated by subjects who participated in the focus group discussion.
Internet Gaming Addiction
To assess risk of internet gaming addiction, we used the Online Game Addiction Scale for Adolescents developed by the Korean Agency for Digital Opportunity and Promotion . This scale is currently used to screen for IGA among adolescents in Korea. The scale is a 20-item self-report measure; each item is rated on a 4-point Likert scale ranging from 1=“not at all” to 4=“always.” The total score for the scale ranges from 20 to 80, with higher overall scores indicating greater risk of IGA. Cronbach’s alpha in the current study was 0.945, indicating high internal consistency.
Peripheral venous blood samples for plasma NE and serum cortisol assays were drawn from participants by professional nurses following standard laboratory procedures for assays. For each subject, 5 milliliters (mL) of venous blood was extracted using a heparin anticoagulation vacuum tube. Levels of plasma NE were measured by high-performance liquid chromatography (HPLC, Agilent 1200 series, Agilent Technology, USA). Serum cortisol levels were analyzed by chemiluminescent immunoassay using ADVIA Centaur and ADVIA Centaur XP systems (ADVIA Centaur XP, Siemens, USA). The ADVIA Centaur cortisol assay is a competitive immunoassay using direct chemiluminescent technology.
Statistical analysis was performed to examine the relationship of gaming reasons to IGA risk and biological indicators. Data were analyzed using IBM SPSS statistics ver. 20.0 (IBM Co., Armonk, NY, USA). Descriptive statistics such as frequency, percentage, mean, and standard deviation were used to summarize the subjects’ demographic and internet gaming-related characteristics. ANOVA was used to compare differences in levels of plasma NE and serum cortisol and IGA risk based on the four categories of internet gaming reasons, with Scheffe post-hoc tests. Analyses of categorical variables by the four gaming reason groups were analyzed using χ2-tests. A p-value of <.05 was considered statistically significant.