Systematized scoping review of instruments for measuring adolescent digital technology use

Review protocol

This scoping review was conducted in accordance with the PRISMA-ScR (Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews) guidelines.¹² We followed established frameworks for scoping reviews,¹³ which emphasise systematically mapping the literature to identify key concepts, instruments, and gaps. The review protocol was developed a priori and made publicly available via the Open Science Framework (OSF id: eb4sfv, available at OSF Registries | Working title: Systematic Scoping Review of tools and instruments for measuring adolescent digital technology use).

Search strategy

A search was conducted across three comprehensive electronic databases: PubMed, Web of Science (WoS), and Scopus. These databases were chosen for their extensive coverage of biomedical, behavioural sciences, and multidisciplinary research, effectively capturing studies relevant to adolescent digital technology use. These databases were searched from January 2014 through the end of 2025 (the date of the last search was December 14, 2025), ensuring broad coverage over the period of widespread digital technology adoption relevant to contemporary digital technology users.

The search strategy combined keywords and controlled vocabulary (i.e. MeSH terms) using Boolean operators to represent three key concepts: the target population (adolescents), the phenomenon of interest (digital technology/device use), and measurement instruments:

TITLE (adolescen∗ OR teen∗ OR youth OR “young adult∗” OR juvenil∗ OR “school-age∗”) AND (“digital device∗” OR “digital technolog∗” OR “mobile device∗” OR “screen time” OR smartphone∗ OR internet OR “social media” OR gaming OR “video game∗” OR computer∗ OR tablet∗ OR television∗ OR wearable∗) AND (measure∗ OR scale∗ OR questionnaire∗ OR instrument∗ OR survey∗ OR assessment∗).

These terms were adjusted for each database's syntax and subject headings. We applied filters to limit search results to peer-reviewed articles published in English. In line with the registered protocol, the search was applied to the title field to prioritise precision and feasibility within a high-volume literature; we acknowledge that this, together with the streamlined keyword set, may have reduced sensitivity and missed some eligible instruments.

Eligibility criteria

Studies were included based on predefined inclusion and exclusion criteria, following the Population–Concept–Context framework commonly used in scoping reviews.¹³ These criteria were established a priori in the protocol to guide screening decisions (OSF id eb4sfv). The target population was adolescents aged 10–19 years, though we also included studies involving broader age ranges (e.g., 8–20 years) if adolescent data or relevant instruments were identifiable. Eligible device types included phones, tablets, computers, gaming consoles, e-readers, televisions, and wearables.

We included studies examining a range of purposes of use, including interaction, communication, informal learning, information searching or browsing, entertainment, content sharing or posting, purchasing, and creating. We excluded formal digital learning activities and e-health interventions directed by a health provider. Study designs eligible for inclusion were primary research studies employing measurement tools, instruments, surveys, or other data collection approaches that quantified digital technology use. This included validated or ad-hoc self-report questionnaires, surveys, or scales (e.g., frequency of smartphone use, duration of social media use, Internet addiction scales), as well as objective recording instruments or diaries (e.g., screen time tracking apps or device logs). We included all geographies and all English-language, peer-reviewed published literature, and excluded editorials, commentaries, and opinion pieces that did not contain original data.

Study selection process

The study selection proceeded in two stages, following the predefined inclusion/exclusion criteria. First, two reviewers independently screened titles/abstracts for potential relevance. Next, two reviewers independently assessed the full-text articles to determine final inclusion. Any disagreements regarding inclusion were resolved through discussion between the two reviewers or by consultation with a third reviewer. To ensure maximum consistency and mitigate screening bias, independent evaluations from both reviewers were applied iteratively to all articles. Initial disagreements regarding inclusion were systematically resolved through collaborative discussion and consensus-building between the two reviewers, or by consultation with a third reviewer when necessary. While formal pre-resolution reliability coefficients were not archived due to this iterative consensus workflow, full alignment was reached on all finalised inclusions. The study selection process was systematically documented and summarised in a PRISMA flow diagram presented in the Results section (Fig. 1).

Data extraction

To make sense of the data, we developed an analytical framework adapted from logic model approaches,¹⁴ to characterise the experience of digital technology use and related experiences across inputs, processes, outputs, outcomes, and impact¹⁵) (Fig. 2) Similar conceptual logic models have been successfully utilised in other non-interventional scoping reviews to systematically map and categorise complex qualitative methodologies, user experience domains, and technology classifications for example.^16,17 Consistent with this conceptual framework, the review considered not only direct behavioural measures of digital technology use but also constructs explicitly linked to digital technology use and situated within the broader pathway through which use is shaped, experienced, and manifested. Inputs refer to antecedent conditions, motivations, contextual factors, and enabling characteristics that shape digital technology use; processes refer to the behavioural patterns and characteristics of use itself; outputs represent proximal experiences, exposures, or immediate consequences associated with use; outcomes represent broader behavioural, social, functional, educational, or psychosocial consequences; and impacts represent more distal effects on health, development, education, or wellbeing. The framework was used to guide data extraction, categorisation and synthesis of identified instruments.

Data extraction occurred in two stages. The initial extraction consisted of a standardised form capturing basic details of included studies (study title; year; authors; source-title; description of instrument used; and whether it relies on a recognised scale/validated instrument or on self-developed/ad-hoc questions). From this, a comprehensive list of identified measurement tools was generated. Next, for each unique tool identified, a secondary targeted search was conducted to locate original studies describing development, validation, and psychometric testing.

Aligned to the framework described in Fig. 2, information was extracted as part of the second stage and included: Name of Instrument and Acronym; Language(s) of Instrument; Target Respondents/Sample (age range, demographics); Informant (self-report, proxy-report, or passive collection); Administration Method (paper-based, digital/app-based); Administration Setting (home, school, clinical); Types of technology/device used; Dimensions of Use Measured including frequency, duration, objectives of engagement, patterns of engagement, or outcomes); Number of Items in the tool; Psychometric Properties (validity, reliability metrics such as Cronbach's alpha, test-retest reliability, sensitivity, specificity, etc.). Extracted data were compiled into structured tables, reviewed independently by two reviewers for completeness and accuracy. The final data extraction table formed the basis for our analysis of instruments in the results.

Quality and bias assessment

Although scoping reviews traditionally do not require quality appraisal, we included an assessment of the psychometric quality of each measurement tool to enhance the practical applicability of our findings. The COSMIN (Consensus-based Standards for the Selection of Health Measurement Instruments) guidelines were adapted for the purpose of this scoping review to allow for a broad descriptive appraisal rather than a full methodological evaluation.¹⁸ Specifically, we focused on key psychometric domains most relevant to instruments assessing behavioural constructs in adolescents—content validity, construct validity, internal consistency, and test–retest reliability. For each domain, instruments were appraised against the standard COSMIN definitions¹⁸: content validity (relevance, comprehensiveness, and comprehensibility of items, including evidence of target-population or expert input during development), structural and construct validity (e.g., factor-analytic structure and tests of hypothesised relationships), internal consistency (e.g., Cronbach's alpha or McDonald's omega), and test–retest reliability (temporal stability, e.g., intraclass correlation). These features were documented for completeness but were not required for inclusion, given the scoping review's objective of mapping existing tools rather than formally rating or ranking them. Each measurement tool was rated on its reliability and validity based on the standard COSMIN rating scale (inadequate, doubtful, adequate, good). Risk of bias, such as conflicts of interest or inadequate methodological detail, was documented narratively. Instruments were given an overall score of strong, moderate, or poor quality based on these criteria. Consistent with scoping-review methodology, this appraisal was descriptive only: no instruments were excluded on the basis of psychometric quality, and all were retained to map the measurement landscape rather than to judge which instruments are acceptable for use.

Data synthesis

Data were synthesised using a structured narrative approach comprising several areas. Identified measurement tools were categorised based on key characteristics including the type of technology or device assessed (e.g., smartphones, tablets, gaming consoles, wearables), primary dimensions measured (frequency, duration, purpose, engagement patterns, context), and developmental specificity (instruments specifically designed or adapted for adolescents). Common themes and recurring patterns within the literature were identified, such as prevalent areas of measurement (addictive behaviours, mental health) and developmental sensitivity of instruments. Instruments measuring similar outcomes were compared to highlight areas of convergence, divergence, and significant gaps.

Deviations from the registered protocol

The review followed the registered protocol (OSF: eb4sfv) with a few deviations. First, the search window was extended from the registered end date of December 2024 to December 2025 (final search 14 December 2025) to capture the most recent literature. Second, the search string was streamlined: population terms were focused on adolescence (the broader terms child∗ and underage∗ were not retained), overlapping device and media terms were consolidated, and the registered restriction to a time- or use-related term (time OR use) was not applied, which broadened retrieval to instruments assessing engagement, attitudes, and competencies alongside use. Third, the registered quality rubric (rating instruments as strong, moderate, or poor) was operationalised using an adapted COSMIN framework, with domain-level appraisal of content validity, construct validity, internal consistency, and test–retest reliability informing the overall rating. These deviations did not change the review's eligibility criteria or objectives.

Role of funding sources

None.

Study selection process

The database search identified a total of 1600 records. After removing 711 duplicates, 889 titles/abstracts were retained for screening. Title and abstract screening excluded 306 records that did not meet the inclusion criteria (e.g., not relevant based on criteria, described in the methods, non-English publication, or published before 2014), leaving 583 articles for full-text review. Of these, 50 were excluded at the full-text stage because they were not published in English (n = 8) or were conference abstracts, reviews, or other studies not relevant to the review (n = 42), resulting in 533 studies being included in the final scoping review. The study selection process is summarised in the PRISMA flow diagram (Fig. 2).

Overview of included studies

The 533 included studies were published between 2014 and 2025, with a notable increase in publications over time (Table S1). Only 120 studies appeared before 2020, whereas the remainder were published in the last five years (2020–2025), reflecting growing research interest in adolescent digital technology use, and possibly the increased demand during and after the COVID pandemic. Studies spanned 20+ countries, mainly East Asia, North America, and Europe, with notable underrepresentation from Africa and Latin America. Approximately a third (35%) of studies presented data from Low or Middle Income Countries (LMICs). The included studies were predominantly cross-sectional surveys of adolescent samples drawn from school or community settings.

Categorisation of measurement tools

Across the 533 studies, a wide range of measurement tools were used to assess adolescent digital technology use. We categorised these tools into two broad types: validated instruments (established questionnaires or scales with published validation) and self-developed or ad hoc measures created by authors for a specific study. Just over half of the studies (n = 316, ∼59%) employed one or more validated instruments to measure digital technology use. In some cases, instrument development and validation was undertaken as part of the study itself. The remaining 217 studies (∼41%) relied on study-specific or ad-hoc single-item metrics without formal validation. A frequent approach was to ask adolescents to report frequency or duration of use (e.g., “How many hours per day do you spend on social media?”), or to record screen time through diaries or device logs.

A total of 135 validated instruments were identified across the 533 studies. Most of these were brief and easy to administer self-report measures (ranging from 4 to 51 items). Only 2 alternative assessment methods were identified, involving passive data collection using a mobile app. Table S2 provides a summary of the key characteristics of these instruments, including the language of the instrument, target respondents, informant, administration, device and features measured, number of items, and key references describing development, validation, and psychometric properties. Most common instrument languages were Chinese, Korean, Spanish and French.

Dimensions assessed

The included studies assessed a range of features relating to digital technology use, with nearly all capturing some aspect of usage frequency or duration—such as daily screen hours, number of logins, or behavioural counts (e.g., messages sent, videos watched). These basic quantitative metrics dominated the literature, reflecting a prominent focus on how often and how long adolescents use technology (Browne et al., 2021). Applying the analytic framework shown in Fig. 2, instruments were mapped across the framework's components to characterise the domains of digital technology use and related constructs represented in the literature (Table 1.) Many of the instruments-based measures went beyond raw frequency to distinguish between different platforms (e.g., social media, gaming, streaming) and specific activity types. Domain-specific instruments, such as the Bergen Social Media Addiction Scale (BSMAS) and Gaming Disorder scales, focused on social media or video gaming, respectively. Other instruments covered multiple use cases, asking separately about time spent on social media, video games, streaming video, or web browsing. Multidimensional tools (e.g., Internet Gratification Scale for Adolescents, Content-based Media Exposure Scale [C-ME]) assessed patterns of engagement, including habitual checking, multitasking, or content types, rather than focussing on specific devices or media types.

Social or contextual factors were rarely included. Few instruments captured situational elements (e.g., device use alone versus with others, at night, or during meals), parental involvement (e.g., rules, supervision), or motivational drivers (e.g., entertainment, escapism, social connection). A small subset of tools implicitly assessed motivation through items addressing boredom relief, mood regulation, entertainment or self-presentation (e.g., the Psycho-Social Aspects of Facebook Use [PSAFU], the Motives for Online Gaming Questionnaire [MOGQ], and the Motivations for Social Media Use [MSMU]). Very few instruments included psychosocial correlates, such as anxiety when unable to access devices, or interference with sleep, academics, or social relationships—typically within tools targeting problematic use. Positive use dimensions, such as enjoyment, peer bonding, or skill-building, were rarely captured.

Notably, over two-thirds of the identified instruments focused explicitly on pathological or problematic use, or addiction (criteria included for e.g., withdrawal, tolerance, dependence, loss of control). These instruments were designed to identify excessive, uncontrolled, or harmful use and mapped onto diagnostic frameworks (e.g., Internet Addiction Test, Social Media Disorder Scale, Smartphone Addiction Scale, etc.). In contrast, less than 2% of the identified instruments explicitly assessed positive dimensions of digital technology use (e.g., eHealth Literacy & Cyber Virtues scales). Constructive behaviours like online learning or creativity were occasionally included, but typically as isolated items rather than dedicated scales.

In summary, available instruments are skewed heavily toward capturing the quantity and patterns of problematic use, whereas nuanced aspects of how and why adolescents engage with technology are captured less frequently by available instruments.

Developmental appropriateness and contextual relevance

An important consideration in digital technology use measurement is whether the available instruments account for adolescent's unique developmental context. We found that most instruments were age-insensitive in nature, rather than adolescent-focused. We considered an instrument age-insensitive if it had not been developed or validated specifically for the target adolescent age range, contained no developmentally specific content (e.g., schoolwork, parental mediation, or peer context) or age-referenced norms, and had been applied to adolescents without developmental adaptation. Most tools had been developed for broad populations (often including adults) or for “youth” in general. Very few tools, such as the Adolescent Pathological Internet Use Scale, included age-relevant content (e.g., schoolwork impairment), the majority lacked developmental specificity. Generic wording was common (e.g., “neglecting responsibilities”), and few tools addressed salient adolescent contexts such as parental dependence, peer influence, or school environments. Only a small subset of instruments (12%), such as the Social Media Disorder Scale, were developed and validated specifically for adolescents (aged 10–17 years¹⁹). Many widely used instruments (e.g., Internet Addiction Test; Young et al., 1998) were adapted from adult versions with minimal modifications (e.g., simplifying language or removing adult-centric items), and few incorporated adolescent input during formative qualitative work.

Although many instruments have been translated into multiple languages and applied internationally (e.g., the Social Media Disorder Scale [SMDS] and the Excessive Internet Use Scale [EIUS]), cross-cultural adaptation was also rare.^20,21 Most studies reported translation but not content modification or validation across cultural groups. Socioeconomic contexts—such as device access, digital norms, or urban–rural differences—were also infrequently considered. As a result, while instruments are widely used across diverse groups and global settings, few are tailored to the unique developmental and sociocultural realities of adolescents.

Quality and bias assessment

A systematic evaluation of validated instruments using adapted COSMIN guidelines¹⁸ revealed substantial variability in psychometric quality (Table S3). Among the identified measurement tools, a significant proportion demonstrated moderate psychometric robustness, with reliability metrics, particularly internal consistency, commonly reported and generally satisfactory. However, validity metrics were inconsistently reported and frequently limited to exploratory factor analysis or simple correlational approaches. Additionally, independent validation studies were scarce, and a considerable number of instruments lacked rigorous psychometric testing beyond initial pilot or exploratory phases. Interpretability and responsiveness were generally not reported as psychometric properties, highlighting critical gaps in current instrument development and validation.

Instruments rated as “strong” (n = 19; 14%) in overall quality were typically those with multiple validations, comprehensive psychometric documentation, and transparent reporting of methodology and limitations. Conversely, instruments categorised as “poor” (n = 14; 10%) often exhibited unclear methodological reporting, limited psychometric testing, or potential conflicts of interest due to validation primarily conducted by the original instrument developers. The majority of instruments were of “moderate” quality (76%).

Outstanding questions

To advance measurement in adolescent digital technology use, several targeted recommendations emerge from this review:

• 1.

  Develop holistic measurement instruments: Accurately capturing adolescents' digital technology use requires attention to a broader range of constructs, including not only inputs such as time and motivation, but also the contexts, meanings, and potential benefits of use. Future research should refine existing conceptual approaches with input from adolescents, parents, and educators, and develop and validate tools that capture positive, normative, and qualitative dimensions of digital use, including digital literacy, creativity, social connection, resilience, and contextual features of technology use. Such efforts would help align measurement practices with the recognition that digital technology use shapes adolescent development in ways that extend beyond risk alone.

• 2.

  Establish Measurement Standardisation: We also advocate for further consensus on core constructs and metrics relating to digital technology use. Standardised definitions and measurement criteria (e.g., “screen time” or definitions of problematic use) would facilitate cross-study comparability and meta-analytic synthesis.

• 3.

  Improve and Diversify Validation Efforts: Rigorous validation studies are needed to examine measurement invariance across different demographic groups, familial and cultural contexts, and developmental stages. Ensuring robust psychometric properties—including responsiveness and interpretability—is critical for tools designed to detect change over time or inform clinical practice.

• 4.

  Incorporate Multi-Method Approaches: Integrating objective usage data (e.g., passive smartphone logging, wearable sensors, ecological momentary assessments) with traditional self-report tools can significantly enhance measurement accuracy and reliability. Hybrid measurement approaches can reduce recall bias and provide richer, contextually detailed data.

• 5.

  Engage Adolescents in Measurement Design: Employing participatory design methods involving adolescents directly in instrument development can improve relevance, face validity, and developmental appropriateness of measurement tools. Adolescents' insights can help identify overlooked aspects of digital experiences and ensure that tools reflect contemporary usage patterns accurately. Researchers should be cautious about clinging to legacy measures for ease of comparability at the expense of relevance.

In summary, the findings from this scoping review highlight the need for comprehensive, balanced, and methodologically rigorous approaches to measure adolescent digital technology use. Improved measurement is not an end in itself but a means to better understand and positively influence adolescent development in the digital age. Addressing identified gaps and adopting recommended strategies will yield more precise, developmentally appropriate, and actionable measurement of adolescent digital technology, facilitating better research, policy-making, and practical interventions in this rapidly evolving field.