New scale validates parental reports of gaming disorder in young children

Scientists have developed and validated the first psychometric instrument designed specifically to assess gaming disorder in children as young as five, filling a conspicuous gap in a research field whose existing measurement tools have, until now, excluded the very youngest and most vulnerable gamers. The new instrument, known as the Parent Proxy Measure for Gaming Disorder Among Young Children (PPM-GD-YC), was created by a multinational team led by Zainab Alimoradi, Mehran Alijanzadeh, and Amir H. Pakpour of Qazvin University of Medical Sciences in Iran, together with Mark D. Griffiths of Nottingham Trent University in the United Kingdom and colleagues from the Deputy of Health at Qazvin University of Medical Science. Published in Discover Psychology, the study addresses a problem that has grown increasingly urgent as touchscreens and mobile games reach children long before they are developmentally capable of reporting on their own behavior. Because self-report questionnaires require reading comprehension, metacognitive insight, and sustained attention that children aged five to seven typically lack, researchers have long been unable to quantify problematic gaming patterns in this age group. The new tool sidesteps that limitation by asking the people who observe young children most closely, their parents and guardians, to rate the addiction-relevant features of their children’s gaming behavior.

The theoretical foundation of the PPM-GD-YC is the components model of addiction, a widely cited framework developed by Griffiths that holds that all addictive behaviors, whether chemical or behavioral, share a set of core components. Accordingly, the preliminary version of the scale contained eleven items designed to capture six such components: salience, meaning the degree to which gaming dominates the child’s thinking and behavior; conflict, referring to interference with other activities or with relationships; tolerance, the need for increasing amounts of gaming to achieve the same effect; relapse, the failure to reduce gaming despite repeated attempts; withdrawal, the negative emotional states that emerge when gaming is stopped; and mood modification, the use of gaming to change emotional states. This framework aligns conceptually with the clinical descriptions of gaming disorder found in the World Health Organization’s International Classification of Diseases, in which impaired control, increasing priority given to gaming, and continuation despite negative consequences are the defining features. By translating these constructs into items that a parent can realistically observe and rate in a preschool-aged or early school-aged child, the team created something the field has never had: a structured, theory-driven proxy measure for the youngest gamers.

The psychometric evaluation was conducted between June and November 2023 in Qazvin province, Iran, using a two-stage quota random sampling design intended to produce a sample representative of the province’s parent population. In total, 5,630 parents completed the survey, with fathers averaging 39.0 years of age and mothers 34.2 years. Parents reported that their children spent, on average, one and a half hours per day playing video games, a figure that underscores how deeply embedded gaming already is in early childhood routines. The scale development followed a rigorous multi-step analytical pipeline combining classical and modern psychometric approaches. First, exploratory factor analysis (EFA) was used to uncover the latent structure underlying the eleven candidate items. Second, confirmatory factor analysis (CFA) tested whether the structure suggested by the EFA held in the data according to pre-specified fit criteria. Third, Rasch analysis, an item-response-theory technique that evaluates whether individual items contribute meaningfully to measurement along a single continuum, was applied to refine the instrument further. This triangulation of methods is considered best practice in scale validation because each approach interrogates different aspects of measurement quality.

The analytical winnowing process was substantial. Of the eleven initial items, six were retained in the final version of the scale, suggesting that several candidate items were either redundant with others, poorly matched to the underlying construct, or psychometrically weak in this young-child population. The surviving six-item measure demonstrated a unidimensional factor structure, meaning that all retained items tap a single underlying dimension of gaming-related impairment rather than multiple separate facets. This unidimensionality was supported consistently across EFA, CFA, and Rasch analysis. The confirmatory factor analysis produced fit indices that comfortably exceeded conventional thresholds: the comparative fit index (CFI) reached 0.976 and the Tucker-Lewis index (TLI) 0.959, both above the commonly accepted 0.95 benchmark, while the root mean square error of approximation (RMSEA) of 0.06 and the standardized root mean square residual (SRMR) of 0.048 fell within the ranges generally interpreted as indicating acceptable to good model fit. In practical terms, these statistics indicate that the six-item structure provides an accurate and parsimonious representation of how parents’ responses co-vary.

Rasch analysis provided additional granularity about the quality of individual items. The infit mean-square statistics ranged from 0.91 to 1.16 and the outfit mean-square statistics from 0.90 to 1.12, both falling within the approximate 0.7 to 1.3 window that Rasch methodologists regard as productive measurement. Mean-square values near 1.0 indicate that items perform as expected under the Rasch model, neither overfitting the model by being predictable to the point of redundancy nor underfitting by introducing unpredictable noise. Internal consistency, measured by Cronbach’s alpha, was 0.759, which sits within the acceptable range for a brief screening instrument, though it is naturally more modest than one would expect from longer clinical scales. For a six-item proxy measure intended for population-level use, this level of homogeneity among items represents a reasonable trade-off between brevity and reliability, and the authors emphasize that the brevity is itself a feature, since short parent-reported instruments are far more feasible to deploy in large epidemiological surveys than lengthy clinical batteries.

However, the study is notably candid about an important limitation. Person separation reliability, a Rasch-derived index that reflects the instrument’s ability to distinguish between individuals along the measured continuum, was modest at 0.57. In Rasch terms, this means the scale can reliably differentiate only about two distinct levels of the trait across the sample, which is adequate for comparing groups but insufficient for confident measurement of any single child. The researchers are explicit that the PPM-GD-YC should therefore be used exclusively for group-level comparisons, such as in epidemiological research and large-scale screening studies, and should not be employed for individual clinical diagnosis. They caution further that individual scores should be interpreted with extreme caution, and they note that the construct validity evidence presented remains preliminary because discriminant validity, known-groups validity, and criterion validity have not yet been established. In other words, it remains to be demonstrated that the scale distinguishes gaming disorder from related constructs, that it differentiates known clinical groups, and that its scores correspond to diagnoses made through structured clinical interviews. The authors also recommend that future research prioritize developing a longer version of the measure to improve person separation reliability and, crucially, to establish validated cut-off scores against clinical gold standards.

The significance of the new instrument lies partly in what it reveals about the demographics of early exposure. Existing self-report measures of gaming disorder are validated only for children and adolescents aged roughly 9 to 18 years, which has left researchers blind to the earliest window of gaming adoption precisely during the developmental period when self-regulation capacities are still maturing and when habits, both adaptive and maladaptive, take root. With children as young as five now regularly engaging with smartphones and tablets, a measurement vacuum has existed at exactly the point in development where early intervention would arguably have the greatest preventive value. A parent proxy approach, in which a caregiver who observes the child’s daily behavior completes the ratings, is the methodological standard for assessing young children in fields such as pediatric quality-of-life research and developmental psychopathology, and its application to behavioral addiction represents a natural but previously unexecuted extension of that tradition.

The research also carries practical implications for public health surveillance. Because the PPM-GD-YC is brief, parent-completed, and psychometrically anchored to a single underlying continuum, it is well suited to incorporation into birth cohorts, school-based surveys, and national health studies where questionnaire burden must be minimized. Group-level comparisons enabled by the tool could, for example, track how average problematic gaming scores vary with daily gaming hours, socioeconomic factors, or parental screen habits across communities, generating the epidemiological evidence base needed to inform guidance for families and policymakers. The finding that the average young child in the sample gamed for ninety minutes daily suggests that such surveillance is timely, since even moderate daily exposure in early childhood may shape the trajectory of later digital media use.

The study was approved by the Medical Research Ethics Committee at Qazvin University of Medical Sciences (IR.QUMS.REC.1401.140) and conducted in accordance with the Declaration of Helsinki, with written informed consent obtained from all participating parents. The authors declare that no financial support was received for the conduct of the study, although one co-author, Mark D. Griffiths, discloses research funding and consultancy relationships within the gambling research field unrelated to the present work. The article, published as an open-access paper with a permanent DOI, will be of interest to developmental psychologists, addiction researchers, pediatricians, and anyone concerned with the growing role of interactive screens in early childhood. As the first validated proxy measure of its kind for children aged five to seven, the PPM-GD-YC opens a previously closed observational window, even as its authors firmly warn that the tool’s proper role for now is population-level screening rather than the diagnosis of any individual child.