Population and Sample
The target population of this current study was youth aged 15–24 years in Thailand. The sample size was calculated using the infinite population proportion formula, in which Proportion (P) = 0.205 (20.50% is a rate of preoccupied and likely addicted online game playing [13]), Error = 0.0205 (10% of P), Alpha = 0.05, Z (0.975) = 1.96, and Cluster effect = 2. A total of 2,980 people were estimated, and 15% of that number was added to prevent non-responses and incomplete responses. The final sample size was thus 3,427 people.
A stratified two-stage cluster sampling method was conducted to select representative samples of the study population in the northern region of Thailand. Firstly, a stratified random sampling approach was used to classify the region into three landscapes: the upper North, central North, and lower North. Secondly, a cluster random sampling method was applied to draw two provinces from each stratum, yielding a total of six provinces. Thirdly, a cluster random sampling method was employed again to select one secondary school, one vocational school, and one university from each province, yielding a total of 18 institutions. Lastly, representative classes were drawn from each institution by using a simple random sampling method (lottery method). Lists of student names from each classroom were then gathered (the randomization of the classrooms was processed until the required samples were obtained). All procedures were approved by the Naresuan University Institutional Review Board. All participants were informed about the study, and they signed the consent form prior to participation.
Measurements
A self-report questionnaire was used to collect quantitative data. This research tool consisted of five parts as follows:
Part 1: Demographic and general information - checklist and short-answer questions were used to assess gender, age, educational level, and latest Grade Point Average (GPA).
Part 2: Mobile Game Addiction Screening test - GAST (Game Addiction Screening Test for Adolescents) [14], a Thailand standard game addiction test, was applied. This test comprised a total of 16 questions. The word “mobile game” instead of “game” was substituted to the original test for specific purposes of this research. The examples of this test are “I so much play mobile games that I am much less interested or do other activities”, “I often play mobile games until forgetting the time”, “since I have played mobile games, my relationship with family has deteriorated”, “I used to play mobile games so late that I can't wake up to go to school”. All questions were rated on a 4-point Likert scale: “Not at all”, “Probably not”, “Probably Yes”, and “Definitely yes”. The interpreting criteria was classified into three levels as “normal or non-problematic game playing”, ‘preoccupied or initially problematic game playing, and “likely addicted or considerably problematic game playing”.
Part 3: Depression Scale - the Thai version of the Patient Health Questionnaire for Adolescents: PHQ-A [15] was adopted to assess the depression levels of the participants. This scale consisted of nine items, for example, “You feel depressed, down, irritable, hopeless, bored, not interested or not enjoyed when doing any activities”, “You have trouble falling, staying asleep all day or sleeping too much”, “You have a poor appetite, weight loss, or overeating”. All questions were rated on a 4-point Linkert scale: “Not at all”, “Some days”, “Frequent”, and “Daily”.
Part 4: Loneliness Scale – the Thai version of the loneliness scale [16], developed from the original 6-Item Revised UCLA Loneliness Scale (RULS-6) [17], comprising of 6 questions, was employed. The example questions include “How often do you feel lacking companionship?”, “How often do you feel alone?”, “How often do you feel no longer close to anyone?”, “How often do you feel left out?” All questions were rated on a 4-point Likert scale: “Often”, “Sometimes”, “Rarely”, and “Never”.
Part 5: Social anxiety scale - Social Anxiety Scale among Adolescents (La Greca & Lopez, 1998) was translated into the Thai version and checked for translation quality by language experts. This scale included 10 questions such as “I am worried about what other people say about me”, “I am worried that other people won't like me”, “I am worried about what other people think of me”, “I feel like people are mocking me”. All questions were rated on a 3-point Likert scale: “Regular”, “Sometimes”, and “Never”.
Content validity was evaluated using the Index of an Item Objective Congruence (IOC) procedure by three experts. The IOC value was between 0.67 -1.00. The questionnaire parts 2–5 were then tried out with 30 non-sample youths to test its reliability. As a result, Part 2, Part 3, Part 4, and Part 5 had Cronbach’ s alpha coefficient of 0.903, 0.867, 0.867, and 0.867, respectively.
Data collection
Participants were explained the research's importance, necessity, and objectives during the data collection process. The researchers distributed copies of the questionnaire to the participants after they obtained consent (Consent Form) to participate in the study (this document was sent to the parents of participants under 20 years to ensure that a signature was given). The researchers explained how to answer the questionnaire in detail, as well as clarifying how their confidentiality would be maintained, including 1) the need not specify a name, 2) the data collected will not be shared with teachers and parents, and 3) an overview of the research results will be presented anonymously. It took a maximum of 90 minutes to complete the questionnaire. After finishing all processes of data collection, 3,424 copies of the questionnaire were completed and returned with a response rate of 99.9%.
Data analysis
The quantitative data were analyzed using SPSS a statistical package, with descriptive statistics, including number, percentage, mean and standard deviation (SD). Amos 22.0 was used to analyze statistical Structural Equation Modeling (SEM) to test the relationship between mobile game addiction and depression, loneliness, social anxiety, and academic performance (latest Grade Point Average (GPA) both overall and classified by gender and educational levels. Confidence Interval (CI) was set at 95%.