Reliability Generalization Meta-Analysis of Internet Gaming Disorder Scale

Area of Science:

• Psychological assessment
• Psychometrics
• Behavioral sciences

Background:

• Internet Gaming Disorder (IGD) is a growing concern.
• Accurate measurement of IGD is crucial for research and clinical practice.
• The Internet Gaming Disorder Scale (IGDS) is widely used but its reliability across versions requires examination.

Purpose of the Study:

• To conduct a reliability generalization meta-analysis of the IGDS.
• To assess the psychometric adequacy of four IGDS versions: original, abbreviated, dichotomous, and ordinal.
• To determine which IGDS version is most reliable for diverse applications.

Main Methods:

• Systematic review of 26 studies (2015-June 2022) using the IGDS.
• Reliability generalization meta-analysis using random coefficients (RC) and varying coefficients (VC) models.
• Analysis of Cronbach's alpha across different IGDS versions and response formats.

Main Results:

• Ordinal versions (IGDS9P, IGDS27P) and IGDS27D demonstrated reliability coefficients > 0.80.
• The IGDS9D version showed reliability coefficients > 0.70.
• High heterogeneity (I2 > 95%, Rb > 90%) was observed across all versions, indicating sample-size dependency.

Conclusions:

• Ordinal and longer dichotomous IGDS versions offer acceptable to high reliability.
• Abbreviated versions do not consistently ensure acceptable reliability.
• The IGDS27P is recommended for applications demanding high precision in measuring Internet Gaming Disorder.