Big data in human behavior research: a contextual turn

From behaviorial insights to big data and back againThe widespread adoption and development of digital tracking technologies and Big Data analytics have fundamentally transformed how we study human behaviour (see Favaretto, De Clercq, Schneble, & Elger [13] for a critical discussion on definitions of Big Data).Wearable sensors [26], GPS-enabled devices [14], online activity logs [24], and social media platforms (e.g., Lazer et al.,[20], Ruths & Pfeffer [28]) continuously generate vast amounts of data that capture (real-time) human interactions, mobility patterns, and decision-making processes at an unprecedented scale and granularity.These technologies quantify human actions, turning individual and collective behaviours into computationally analyzable data points.This process is by no means unidirectional-while human behaviour informs data generation, the analysis of Big Data aims to extract insights that, in turn, enhance our understanding of behaviour itself.In practice, once collected, Big Data must be structured, processed, and analyzed to uncover behavioral patterns.In recent years, advances in Artificial Intelligence (AI) and Machine Learning (ML) have made it possible to detect correlations and infer trends in human behaviour across different domains.Various techniques such as Natural Language Processing (NLP) [15], network analysis [27], and predictive and simulation-based modelling [4,6] have significantly enhanced our ability to analyze and interpret large-scale behavioral data.Despite its strengths, Big Data analytics has inherent limitations when attempting to capture and reconstruct the full complexity of human behavior.Issues such as incomplete datasets, erroneous data inputs, inconsistencies in data structures, schema mismatches, and formatting discrepancies continue to hinder data practitioners and affect the accuracy, integration, and interpretability of analyses.As widely acknowledged, AI-driven analysis can reinforce biases in data collection and decision-making if not critically examined through ethical lens (e.g., Liu & Liu [22], O'Neil,[23]).Big Data often struggles to capture cultural, social, and psychological nuances due to a lack of situational understanding (e.g., Sivarajah, Kamal, Irani, & Weerakkody [30]).For instance, algorithmic systems may misclassify applicants by failing to account for