What Drives Identity Development in Adolescent Gamers? A Three-Wave Study of Gaming Motives and Intensity

Introduction Adolescence is a key period for identity development, marked by the task of consolidating a coherent sense of self (Erikson, 1968; Meeus, 2011). Eriksonian work conceptualizes this process through identity synthesis (a coherent, committed sense of self) and identity confusion (uncertainty and doubt). Longitudinal research using the Erikson Psychosocial Stage Inventory (EPSI; Rosenthal et al., 1981) confirms that these are developmental trajectories that predict psychosocial adjustment: higher synthesis and lower confusion are consistently associated with better functioning, while deviations from this normative trend predict disadvantages. This view is essential for understanding how developmental contexts shape the self over time; for example, Reis and Youniss (2004) successfully mapped distinct identity trajectories to social relationships, while Schwartz et al. (2009c) linked increasing confusion trajectories to family functioning. Video games provide a primary context for this identity work (Granic et al., 2020; Muriel, 2022), yet the specific mechanisms linking gaming to identity formation remain understudied. Existing research often relies on cross-sectional designs or focuses narrowly on addiction. For instance, Šporčić and Glavak-Tkalić (2018) found in a cross-sectional sample that low self-concept clarity was strongly linked to escape and coping motives. However, their study focused on predicting problematic gaming (IGD) rather than identity development itself. Similarly, Larrieu et al. (2022) identified distinct motivational profiles in a cross-sectional study, distinguishing between "escapers" (associated with poor psychological health) and "competitive" gamers (associated with healthy functioning), despite both groups reporting high gaming intensity. This leaves a critical gap: we do not know if gaming motives merely reflect existing deficits, or if they actively shape the rate of change (slope) in identity synthesis and confusion over time. Because identity is developmental and these associations are complex, understanding these dynamics requires longitudinal designs that move beyond cross-sectional snapshots to model change (Crocetti et al., 2023). Just as social connection and recreation can support identity consolidation by offering "practice worlds" for role rehearsal (Arbeau et al., 2020; Forsyth et al., 2017), compensatory motives may fuel specific confusion. By applying a longitudinal framework to these dynamics, we can better understand how the reasons adolescents play influence their developmental trajectory. Design and Aim Although cross-sectional studies consistently identify a bifactor structure as the optimal measurement model for the EPSI (Dimitrova et al., 2019; Schwartz et al., 2009a), we decided against using a Longitudinal Bifactor Growth Model for this study. Previous longitudinal research attempting this structure with the EPSI has encountered significant challenges. Notably, Hatano et al. (2022)—despite a massive sample size (N > 5000)—reported that a complex longitudinal model including synthesis and confusion failed to converge, forcing them to revert to separate models. The only successful longitudinal EPSI bifactor study to date (Hatano et al., 2018) was limited to two waves (using Latent Change Scores), which is mathematically simpler than the three-wave growth curve required for our design. Therefore, we will adopt a Parallel Process Latent Growth Curve Model as our primary analytic strategy. This alternative, supported by Verschueren et al. (2018), models identity synthesis and confusion as distinct but correlated developmental trajectories and is well-suited for modeling individual differences in developmental change. We will estimate two separate (univariate) latent growth models as a exploratory analysis. Should the parallel model fail to converge due to multicollinearity or model complexity the separate models are also our contingency plan. Regardless of the final model structure, we examine how individual differences in the growth factors (intercept and slope) are predicted by adolescents’ baseline gaming motives and gaming intensity. In addition, we address a limitation in prior motivational research—which has predominantly relied on male-dominated samples—by explicitly testing whether these associations are similar or different for boys and girls.