Factors influencing consent for the processing of personal data : a privacy-neutral taxonomy for the online world

An online world has emerged in which the collection and processing of personal information has become a foundational element of the digital society. Within this environment, individuals are overwhelmed with consent requests to which they, ostensibly, pay little attention. Problems concern (i) the dynamics of online consent, which is fundamentally ill understood, and (ii) a research narrative in which privacy-concern is pervasive, despite research indicating significant demographic heterogeneity. To provide a mechanism to better understand online consent, a novel conceptual model for informed consent decision making is developed. This original model adds value to our understanding of online consent practice by incorporating the Autonomous Authorisation model of informed consent within a mainstream behavioural model, the Theory of Planned Behaviour. To provide insight into the dynamics of online consent, a privacy-neutral taxonomy of eight variables influencing online consent behaviour, their interrelationships and relative importance are established. The privacy-neutral taxonomy of variables that influence online consent behaviour is constructed via a thematic analysis of extant literature. Their interrelationships and relative importance is determined, using Interpretive Structural Modelling and Interpretive Ranking Process (IRP) techniques respectively. The ranking of variables is validated using Principal Component Analysis (PCA) on a dataset (n=803) collected via a survey of UK-resident Millennials. Regression is performed on the dataset to determine the willingness of consent provision. The eight discerned consent-influencing variables, from most to least important are: situation, emotion/mood/personality, past experience, requester characteristics, resource requirements, cognitive characteristics, trust and actual understanding. A high degree of dependence was found between most of the factors. The PCA validated the IRP-derived ranking and reduced these to four variables: situation, cognition, familiarity and trust. The regression reveals that the willingness of consent provision is a poor predictor of authorisation behaviour and that, as the understanding of the consequences of consent increase, willingness decreases.