The Ethics of Big Data: Current and Foreseeable Issues in Biomedical Contexts

The capacity to collect and analyse data is growing exponentially. Referred to as ‘Big Data’, this scientific, social and technological trend has helped create destabilising amounts of information, which can challenge accepted social and ethical norms. Big Data remains a fuzzy idea, emerging across social, scientific, and business contexts sometimes seemingly related only by the gigantic size of the datasets being considered. As is often the case with the cutting edge of scientific and technological progress, understanding of the ethical implications of Big Data lags behind. In order to bridge such a gap, this article systematically and comprehensively analyses academic literature concerning the ethical implications of Big Data, providing a watershed for future ethical investigations and regulations. Particular attention is paid to biomedical Big Data due to the inherent sensitivity of medical information. By means of a meta-analysis of the literature, a thematic narrative is provided to guide ethicists, data scientists, regulators and other stakeholders through what is already known or hypothesised about the ethical risks of this emerging and innovative phenomenon. Five key areas of concern are identified: (1) informed consent, (2) privacy (including anonymisation and data protection), (3) ownership, (4) epistemology and objectivity, and (5) ‘Big Data Divides’ created between those who have or lack the necessary resources to analyse increasingly large datasets. Critical gaps in the treatment of these themes are identified with suggestions for future research. Six additional areas of concern are then suggested which, although related have not yet attracted extensive debate in the existing literature. It is argued that they will require much closer scrutiny in the immediate future: (6) the dangers of ignoring group-level ethical harms; (7) the importance of epistemology in assessing the ethics of Big Data; (8) the changing nature of fiduciary relationships that become increasingly data saturated; (9) the need to distinguish between ‘academic’ and ‘commercial’ Big Data practices in terms of potential harm to data subjects; (10) future problems with ownership of intellectual property generated from analysis of aggregated datasets; and (11) the difficulty of providing meaningful access rights to individual data subjects that lack necessary resources. Considered together, these eleven themes provide a thorough critical framework to guide ethical assessment and governance of emerging Big Data practices.     

Destination and source memory in Huntington's disease

Destination memory refers to the recall of the destination of previously relayed information, and source memory refers to the recollection of the origin of received information. We compared both memory systems in Huntington's disease (HD) participants. For this, HD participants and healthy adults had to put 12 items in a black or a white box (destination task), and to extract another 12 items from a blue or a red box (source task). Afterwards, they had to decide in which box each item had previously been deposited (destination memory), and from which box each item had previously been extracted (source memory). HD participants showed poorer source as well as destination recall performance than healthy adults in the proposed tasks. Correlation analysis showed that destination recall was significantly correlated with episodic recall in HD participants. Destination memory impairment in HD participants seems to be considerably influenced by their episodic memory performance.     

Effect of previous fighting on the dynamic of agonistic encounters in zebrafish males

Animal Cognition - Aggression among individuals which compete for resources such as food or territory, or to establish dominance relationships, can cause injuries that may be risky for the...     

Role of serotonin in human aggressive behaviour

This study aimed to evaluate a peripheral serotonergic marker, ³H-imipramine (³H-IMI) binding to platelet membranes, in a group of severely aggressive subjects (A), institutionalized since childhood for mental retardation, as compared with suicide attempters (S) and healthy controls (H). The maximum binding capacity of ³H-IMI to platelet membranes was statistically lower in (A) and (S) than in (H). In addition, a statistically significant difference was observed between the Bmax values of aggressive subjects and those of suicide attempters. No changes in the dissociation constant (Kd) of IMI binding were observed. These data provide further supporting evidence for the hypothesis of an abnormality of the 5HT system in aggressive behaviour and suggest that such an abnormality, as reflected by platelet markers, is more severe in suicide attempters. © 1993 Wiley-Liss, Inc.     

Aggression and social experience in domesticated rats.

Small colonies of rats were established, using adult animals that had either received continuous social experience or had been isolated since weaning. Unfamiliar "intruder" rats--with or without postweaning social experience--were exposed individually to the colonies for a 21-hr. period. Behavioral observations and an assessment of the intruder's physical condition indicated that serious fighting, physical injuries, and large weight losses occurred only when an isolation-reared intruder was placed into a colony of socially experienced rats. These results demonstrate that aggression is a joint function of the rearing history of both the colony and the intruder and that social experience plays an important role in the behavioral development of this species.     

Generalized Break Equivalence II: Contextual Control Over A Generalized Pattern of Stimulus Relations

The Psychological Record - In an earlier work (Gomez, Barnes-Holmes, & Luciano, 2001), it was found that although generalized break equivalence was achieved for 3 subjects (i.e., responding...     

The role of values with personal examples in altering the functions of pain: comparison between acceptance-based and cognitive-control-based protocols

The purpose of the present study was twofold. First, to compare the effect of establishing a motivational context of values on pain tolerance, believability, and reported pain, with three experimental conditions: pain acceptance (ACT condition), pain control (CONT condition), or no values (untrained condition). Second, the study aimed to isolate the impact of adding the corresponding coping strategies to both the ACT and the CONT conditions. Thirty adults were randomly assigned to one of the three experimental conditions. The participants went through the pain task in two occasions (Test I and Test II). In Test I, the effects of the ACT-values protocol (which established pain as part of valued action), the CONT-values protocol (which established high pain as opposed to valued action), and the no-values protocol, were compared. In Test II, the effect of adding the corresponding coping strategy to each condition (defusion for ACT vs. suppression for CONT) was examined. Test I showed a clear superiority of the ACT-values protocol in increasing tolerance and lowering pain believability. In Test II, the superiority of the ACT protocol was replicated, while the CONT protocol proved useful to reduce reported pain, in accordance with previous studies.     

Sentences with Strong Psychological Impact in Psychotherapy: Research in Progress

After a brief exposition of Elementary Pragmatic Model about changes in dyadic interactions, the validity and clinical usefulness of the model was evaluated with a newly created test, called SISCI-Sentences. This test is composed of 90 Sentences with strong psychological impact. Administration of this test to non-clinical and clinical participants produced statistically significant differences between the choices of participants in the two groups. Implications of these results for the use of these Sentences in the course of psychotherapy are discussed.