Personal data use is increasingly permeating our everyday life. Informed consent for personal data use is a central instrument for ensuring the protection of personal data. However, current informed consent practices often fail to actually inform data subjects about the use of personal data. This article presents the results of a requirements analysis for informed consent from both a legal and usability perspective, considering the application context of educational assessment. The requirements analysis is based on European Union law and a (...) review of current practices. As the main outcome, the article presents a blueprint which will be the basis for the development of an informed consent template that supports data controllers in establishing an effective and efficient informed consent form. Because the blueprint, and subsequently, the template, distinguishes between legal and usability requirements, it also provides the basis for the mapping of legal requirements in other contexts. (shrink)

BackgroundConsent for data research in acute and critical care is complex as patients become at least temporarily incapacitated or die. Existing guidelines and regulations in the European Union are of limited help and there is a lack of literature about the use of data from this vulnerable group. To aid the creation of a patient-centred framework for responsible data research in the acute setting, we explored views of patients and next-of-kin about the collection, storage, sharing and use of genetic and (...) health-related data for observational research.MethodsWe conducted qualitative interviews (n = 19) with Dutch sudden cardiac arrest survivors who donated clinical and socio-economic data and genetic samples to research. We also interviewed their next-of-kin. Topics were informed by ethics literature and we used scenario-sketches to aid discussion of complex issues.ResultsSudden cardiac arrest survivors displayed limited awareness of their involvement in health data research and of the content of their given consent. We found that preferences regarding disclosure of clinically actionable genetic findings could change over time. When data collection and use were limited to the medical realm, patients trusted researchers to handle data responsibly without concern for privacy or other risks. There was no consensus as to whether deferred consent should be explicitly asked from survivors. If consent is asked, this would ideally be done a few months after the event when cognitive capacities have been regained. Views were divided about the need to obtain proxy consent for research with deceased patients’ data. However, there was general support for the disclosure of potentially relevant post-mortem genetic findings to relatives.ConclusionsSudden cardiac arrest patients’ donation of data for research was grounded in trust in medicine overall, blurring the boundary between research and care. Our findings also highlight questions about the acceptability of a one-time consent and about responsibilities of patients, researchers and ethics committees. Finally, further normative investigation is needed regarding the (continued) use of participants’ data after death, which is of particular importance in this setting. Our findings are thought to be of relevance for other acute and life-threatening illnesses as well. (shrink)

Though rapid collection of big data is ubiquitous across domains, from industry settings to academic contexts, the ethics of big data collection and research are contested. A nexus of data ethics issues is the concept of creep, or repurposing of data for other applications or research beyond the conditions of original collection. Data creep has proven controversial and has prompted concerns about the scope of ethical oversight. Institutional review boards offer little guidance regarding big data, and problematic research can still (...) meet ethical standards. While ethics seem concrete through institutional deployment, I frame ethics as produced. Informed by my ethnographic research at a large public university in the U.S., I explore ethics through two models: ethics as institutional procedures and ethics as acts and intentions. The university where I conducted fieldwork is the development grounds for a predictive model that uses student data to anticipate academic success. While students consent to data collection, the circumstances of consent and the degree to which they are informed are not so apparent, as many data are a product of creep. Drawing from interviews and participant observation with administrators, data scientists, developers, and students, I examine data ethics, from a larger institutional model to everyday enactments related to data creep. After demonstrating the limits of such models, I propose a remodeling of ethics that draws on recent works on data, justice, and refusal to pose generative questions for rethinking ethics in institutional contexts. (shrink)

In the 60+ years that the modern concept of informed consent has been around, researchers in various fields of practice, especially medical ethics, have developed new models to overcome theoretical and practical problems. While (systematic) literature reviews of such models exist within given fields (e.g., genetic screening), this article breaks ground by analyzing academic literature on consent models across fields. Three electronic research databases (Scopus, Google Scholar, and Web of Science) were searched for publications mentioning informed consent models. The titles, (...) abstracts, and if applicable, full publications were screened and coded. The resulting data on fields, models, and themes were then analyzed. We scanned 300 sources from three databases to find 207 uniquely named consent models, and created a network visualization displaying which models occur primarily in one field, and which models overlap between fields. This analysis identifies trends in the consent debate in different fields, as well as common goals of consent models. The most frequently occurring consent models are identified and defined. The analysis contributes toward a cross-disciplinary “consent design toolkit” and highlights that there are more interrelationships between models and fields than are acknowledged in the literature. Where some models are designed to solve distinctively field-specific issues and are specific to biomedical ethics, some may be adaptable and applicable for other fields including engineering and design. (shrink)

In privacy self-management, people are expected to perform cost–benefit analysis on the use of their personal data, and only consent when their subjective benefits outweigh the costs. However, the ubiquitous collection of personal data and Big Data analytics present increasing challenges to successful privacy management. A number of services and research initiatives have proposed similar solutions to provide people with more control over their data by consolidating consent decisions under a single interface. We have named this the ‘consent intermediary’ approach. (...) In this paper, we first identify the eight obstacles to privacy self-management which make cost–benefit analysis conceptually and practically challenging. We then analyse to which extent consent intermediaries can help overcome the obstacles. We argue that simply bringing consent decisions under one interface offers limited help, but that the potential of this approach lies in leveraging the intermediary position to provide aides for privacy management. We find that with suitable tools, some of the more practical obstacles indeed can become solvable, while others remain fundamentally insuperable within the individuated privacy self-management model. Attention should also be paid to how the consent intermediaries may take advantage of the power vested in the intermediary positions between users and other services. (shrink)

Persuasive technologies for health‐related behaviour change give rise to ethical concerns. As of yet, no study has explicitly attended to ethical concerns arising with the design and use of these technologies for vulnerable people. This is striking because these technologies are designed to help people change their attitudes or behaviours, which is particularly valuable for vulnerable people. Vulnerability is a complex concept that is both an ontological condition of our humanity and highly context‐specific. Using the Mackenzie, Rogers and Dodds’ taxonomy (...) of vulnerability, this paper identifies (a) the wrongs or harms to which a person is vulnerable, (b) the source of this vulnerability, and (c) the safeguards needed in response. Two ethical concerns with the design of persuasive technology for vulnerable people are discussed: the concerns of taking into account users' interests and their autonomy. (shrink)

The debate on the right to be forgotten on Google involves the relationship between human information processing and digital processing by algorithms. The specificity of digital memory is not so much its often discussed inability to forget. What distinguishes digital memory is, instead, its ability to process information without understanding. Algorithms only work with data without remembering or forgetting. Merely calculating, algorithms manage to produce significant results not because they operate in an intelligent way, but because they “parasitically” exploit the (...) intelligence, the memory, and the attribution of meaning by human actors. The specificity of algorithmic processing makes it possible to bypass the paradox of remembering to forget, which up to now blocked any human-based forgetting technique. If you decide to forget some memory, the most immediate effect is drawing attention to it, thereby activating remembering. Working differently from human intelligence, however, algorithms can implement, for the first time, the classical insight that it might be possible to reinforce forgetting not by erasing memories but by multiplying them. After discussing several projects on the web which implicitly adopt this approach, the article concludes by raising some deeper problems posed when algorithms use data and metadata to produce information that cannot be attributed to any human being. (shrink)