Modelo psicobioecológico dimensional de phubbing: uma perspectiva tecno-híbrida com machine learning

Abstract

Contemporary technologies have contributed to the development of new behaviors, such as phubbing, which is characterized by snubbing face-to-face social interactions in favor of a smartphone. The academic community has primarily dedicated itself to investigating the predictors of phubbing and its associations with various psychological, behavioral, social, and mental health variables. However, the literature in the field lacks studies that explore data through more contemporary strategies, such as machine learning-based approaches, as well as studies that present an explanatory psychological theory for phubbing behavior. Although a trend has been observed in the literature to view phubbing not merely as a problematic behavior, but as a result of social restructurings derived from the current historical-cultural context, perspectives that treat phubbing as a categorical phenomenon still prevail. In light of this, the present work aims to propose a dimensional psychobioecological model of phubbing from a developmental perspective within a techno-hybrid reality. This framework views phubbing as a behavioral, social, cultural, historical, and bioecological phenomenon, considering personality traits, Instagram addiction, Instagram usage habits, and sociodemographic variables. These factors will be analyzed using machine learning for non-linear prediction and profiling analyses, alongside Structural Equation Modeling. To this end, the thesis is organized into four studies, subdivided into a theoretical section (Studies 1 and 2) and an empirical section (Studies 3 and 4). Study One focuses on updating Bronfenbrenner's Bioecological Theory based on a techno-hybrid perspective of reality, utilizing a seven-step protocol for the development of theoretical models and a narrative literature review. The data sources consisted of scientific articles, books, and book chapters regarding classical BBT and its subsequent updates. Study Two applies the updated version of Bronfenbrenner's Bioecological Theory to analyze phubbing behavior. It addresses the behavior as a phenomenon that can affect or result from contemporary social restructurings, potentially yielding neutral and positive outcomes in addition to negative ones, without being constrained by technological determinism—which can lead to the pathologization of behaviors that may not be pathological, as well as erroneous measurements and interpretations of the relationship between phubbing and other variables. It adopts a multidisciplinary perspective within a reality of hybrid sociability and follows a protocol to operationalize the Process-Person-Context-Time model. Study Three investigated the psychological, behavioral, and contextual mechanisms underlying phubbing through an integrative approach via network analysis, structural equation modeling, and machine learning. Study Four performed profiling of phubbing practitioners using multiple clustering algorithms with distinct assumptions (Partitioning Around Medoids, Density-Based Spatial Clustering of Applications with Noise, and Gaussian Mixture Models). This thesis represents a significant effort to bridge gaps in the literature on phubbing by proposing an updated theoretical framework based on Bronfenbrenner's Bioecological Theory from a techno-hybrid perspective, and by integrating contemporary analytical strategies such as machine learning, thereby offering a more comprehensive and in-depth understanding of the phenomenon. In conclusion, the integration of the studies from the theoretical and empirical sections enables the development of a dimensional psychobioecological model of phubbing behavior, allowing for an understanding of phubbing beyond a categorical, pathological perspective, and framing it instead as part of a hybrid contextual reality undergoing social transformations.