Whereas learning is one of the primary goals of Explainable Artificial Intelligence (XAI), we know little about whether, how, and when explanations enhance users’ learning from feedback provided by Artificial Intelligence (AI). Drawing on Feedback Theory as a fundamental theoretical lens, we formulate a research model wherein explanations enhance informativeness and task performance, contingent on users’ prior knowledge, ultimately leading to a higher learning outcome. This research model is tested in a randomized between-subjects online experiment with 573 participants whose task is to match Google Street View pictures to their city of origin. We find a positive effect of explanations on learning outcome, which is fully mediated by informativeness, for users with less prior knowledge. Furthermore, we find that explanations positively impact users’ task performance, where this effect is direct for more knowledgeable users and fully mediated by informativeness for less knowledgeable users. We seek to elucidate the mechanisms underlying these effects of explanations on learning from AI feedback in focus groups with AI experts and users. By studying the consequences of explanations as part of AI feedback for users in non-routine inference tasks, we advance the understanding of explanations as facilitators of human learning from AI systems.

While considerable literature has addressed the separate use of health information systems, limited research has examined how the use of one system changes following the implementation of another. We address this gap in the literature by developing and testing hypotheses about substitution and complementarity in physicians’ use of health information exchange (HIE), which merges patient data from multiple sources, following the implementation of a local electronic medical record (EMR) system. The hypotheses are anchored in theories foundational to understanding the motivation for information seeking behaviour. The hypotheses are tested with unique data, recorded in system log files, that describe HIE use by physicians in an emergency department before and after EMR implementation. The findings confirm that system use is characterised by substitution when it is defined dichotomously as whether the system is accessed or not (lower HIE access rates after EMR implementation) and by complementarity when it is defined as the extent of system use (higher volume and duration of HIE use after EMR implementation). Our findings suggest that the information seeking literature has underestimated the explanatory value of the cost-benefit approach and the occurrence of complementarity in system use because of its overreliance on a narrow definition of use.

After more than two decades of intensive research, researchers are still struggling to explain the strategic value of information technology (IT) capabilities. The current study suggests that sufficient advancement has not been made in this area of research because of the gap between the richness of theoretical formulations and the minimalism of the reductionistic approach taken to test them. While theoretical formulations describe complex relationships between IT capabilities and competitive advantage, reductionistic explanations rely on models that represent the integration of simple relationships. This area of research can, therefore, benefit from the wider adoption of a holistic approach that attributes competitive advantage to profiles of IT capabilities. This study draws on the resource-based view and configurational theory to develop two reductionistic (direct and mediation) and two holistic (gestalt and profile-deviation) models of IT strategic value. The four models are tested with data collected from IT managers. The empirical analysis illustrates the insights that can be gained by drawing on the relative strengths of reductionistic and holistic explanations of IT strategic value.