Data were collected through a structured online survey administered to a representative sample of 1,043 adolescents residing in Spain. The sampling strategy employed a stratified random design to ensure generalizability to the national population of youth aged 12 to 18 years. The stratification process was conducted in multiple stages. Initially, the sample was segmented proportionally across the seventeen Autonomous Communities of Spain according to population distributions published by the Spanish Statistics Office (Instituto Nacional de Estadística, INE) for 2023. Within each region, municipalities served as primary sampling units, ensuring representation of both metropolitan and rural environments across five population size strata. Final participant selection applied a transversal classification by gender, age, and municipality size. The achieved sample comprised 527 females (50.5%) and 516 males (49.5%), with a mean age of 13.96 years (SD = 1.43). The data collection company ODEC was commissioned to manage fieldwork, which was conducted between April 24 and May 6, 2024. For specific analyses concerning TikTok, a sub-sample of 737 platform users was identified from the full sample (the demographic and structural breakdown of the sample is presented in Table 1).

The study was conducted in strict adherence to ethical standards and in full accordance with the principles stated in the Declaration of Helsinki. The research design received a formal favourable report and ethical approval from the Research Ethics Committee of the Universitat Oberta de Catalunya (UOC), under reference number CE24-PR05, confirming that the research meets national and international guidelines for research on human participants. Given that the research involved minors, a rigorous informed consent and assent process was systematically implemented. In accordance with Organic Law 3/2018 on the Protection of Personal Data and Guarantee of Digital Rights (LOPDGDD), written informed assent was obtained directly from adolescents aged 14 to 17 for data processing, whereas for younger participants aged 12 and 13, written informed consent was obtained from parents or legal guardians prior to survey access. Verbal consent was not used; all participants or guardians provided written documentation before accessing the instrument. The survey instrument included an introductory section detailing the research objectives, the strictly voluntary nature of participation, the right to withdraw at any time without consequence, and the technical measures implemented to guarantee complete anonymity. All data handling by the fieldwork company ODEC and the research team adhered to UNE EN ISO/IEC 27001 standards for information security management. Participants were explicitly informed that their responses would be processed exclusively in aggregate form and used solely for scientific purposes. No individual-level data was retained by the data collection company after the cleaning and anonymisation phase.

The questionnaire was structured into nine blocks designed to capture a holistic view of the adolescent digital experience, aligned with the SMARS project’s theoretical framework. A detailed description of each block follows: •

Block 0 – Socio-demographic profile: Gender (male/female/other), age (12–18 years), autonomous community of residence, municipality size (five strata from <10,000 to >500,000 inhabitants), and current educational level.

Multiple quality control measures were implemented during and after survey administration to ensure the reliability and validity of the dataset. During fieldwork, the online survey incorporated attention-check items designed to detect and exclude respondents providing inconsistent or inattentive responses. Incomplete submissions were automatically discarded by the data collection system prior to export. Following data acquisition, a systematic cleaning and anonymisation protocol was applied. Open-ended responses in ‘other’ category fields were reviewed to remove any incidental personal identifiers and standardised into title case strings where feasible. Ordinal scales were harmonised across questionnaire blocks to ensure coding consistency. Missing values were labelled uniformly. The final dataset includes only fully completed and validated interviews (N = 1,043). Following standard anonymisation procedures applied by ODEC in accordance with ISO/IEC 27001, the dataset was verified to contain no directly or indirectly identifying information before being transferred to the research team.

The technical quality of the dataset was assessed through comprehensive statistical analyses focusing on internal consistency, structural validity, and consistency with established sociological benchmarks. All analyses were conducted using IBM SPSS Statistics version 25.0. Internal consistency of the adapted Social Perception Scale (Block 3, 9 items) was evaluated using Cronbach’s α, yielding a coefficient of 0.89, indicating excellent internal consistency for this cohort.

For the Digital Well-being on TikTok Scale (Block 8, 13 items, administered to N = 737 TikTok users), an Exploratory Factor Analysis (EFA) was conducted using principal axis factoring with oblimin rotation to account for expected inter-factor correlations. The Kaiser-Meyer-Olkin measure of sampling adequacy was satisfactory (KMO = 0.87), and Bartlett’s test of sphericity was significant (p < .001), confirming the suitability of the correlation matrix for factor analysis. The EFA yielded a three-factor solution explaining 60.46% of total variance. This structure differs slightly from the four-factor model proposed in the original instrument by Prakash, ⁶ as the social connection and communion dimensions merged into a single factor for the Spanish adolescent cohort, suggesting cultural specificity in how these constructs are experienced. Cronbach’s α coefficients for all subscales exceeded the threshold of 0.70 conventionally accepted in social science research ( Table 3).

Descriptive and inferential statistical audits were performed to validate expected patterns in the data and ensure consistency with established sociological findings. Regarding gander differences, the data confirm that adolescent girls report significantly more negative perceptions of the impact of digital technologies on their psychological well-being than boys ( Table 4), consistent with prior literature on gander-differentiated vulnerability on visual platforms. ³ Independent samples t-tests further confirmed that TikTok and Instagram users perceive the social impact of digital technologies significantly more positively than non-users on most scale dimensions, validating the social capital value attributed to platform participation within the adolescent peer group.

Analysis of digital well-being dimensions revealed that the Agency subscale (mean = 3.22, SD = 0.91) scored lowest of the three subscales, compared to Social Connection (mean = 3.64, SD = 0.57) and Emotional Resilience (mean = 3.31, SD = 0.60). A one-way analysis of variance (ANOVA) confirmed a significant effect of daily TikTok usage time on agency scores, F(5, 731) = 7.48, p < .05, with post-hoc tests (Tukey HSD) indicating that heavy users (>2 hours per day) reported significantly lower agency than light users. This pattern provides convergent internal validity, demonstrating that self-perceived capacity to set boundaries on TikTok use decreases as usage time increases.