Accountability in Research

Call for Papers
Special Issue on Research Integrity for Data and Computationally Driven Research Fields

Does the newly published guidance on Research Integrity (National Academies of Medicine 2017) effectively address the needs of researchers in areas of emerging technology such as cognitive computing, machine learning, artificial intelligence, computational biology, robotics, and other data-driven and computationally complex research areas? This recent document, among others, reiterates the importance of key principles for good research practice. But, good research practices often emerge only as a field matures. This leaves many researchers in new and cutting-edge fields without clear standards of guidance, exemplars, or even focused discussions to guide their endeavors.
Despite journalistic hype and public clamor over the potential for harm raised by developments in artificial intelligence and related fields, scholars and practitioners in the research administration and research integrity fields have been largely silent on the application of research integrity to this new area. The purpose of this special issue of Accountability in Research is to invite development of positions on the applicability of research integrity standards to these new areas. Questions that might be answered in this special issue include, but are certainly not exhausted by, the following:
● Ought artificial intelligence be credited or acknowledged as an author in published work?
● How ought credit be distributed for authorship in teams working on machine learning or artificial intelligence projects?
● Ought individuals or teams who develop training data sets be credited as authors or contributors to published works that depend on algorithms learning from those data sets?
● If machine learning algorithms assist in the acquisition or sorting of data, what would be sufficient for a full description of these algorithms in data and methods sections?
● How ought the US Common Rule or the EU Directive be modified to provide sufficient guidance for human-computer or human-robot interaction studies?
● How ought animal care and use standards be modified to address emerging animal-computer interaction research?
● How might ethical industry-academic relationships in computationally driven research fields be managed?
● What responsibilities to society do artificial intelligence researchers have that might differ from the responsibilities of those in other fields?
● What existing research governance structures, if any, are best suited to governance of computationally sophisticated research fields?
● Which new concepts ought to be introduced into the study of research ethics in order to account for emerging technologies such as artificial intelligence?
● Which new practices ought to be introduced into the practice of research administration in order to ensure safety and beneficence in artificial intelligence research?

Interested authors should submit an abstract to Dr. Sara Mattingly-Jordan (srjordan@vt.edu) AND Dr. Adil Shamoo (ashamoo@som.umaryland.edu) before or on 1 April 2018. Authors selected for inclusion will be asked to submit a draft of their paper via Editorial Manager (https://www.tandfonline.com/toc/gacr20/current). Draft manuscripts should not exceed 5000 words in length and should not be under consideration in another venue. Papers should be formatted in accordance with the Accountability in Research manuscript submission guidelines. The area of research interest and the question to be answered should be clearly identified in the paper. Training data, data-development or data-engineering processes, and algorithms used should be clearly described in the paper and, when possible, made available to readers.

Inquiries related to this project should be addressed to Dr. Mattingly-Jordan at srjordan@vt.edu.