Progressing digitalization, pervasive mobile computing, and omnipresent social media platforms with billions of users have led to the rise of big data analytics and the implementation of data-driven research approaches in a wide variety of scientific fields. Recently, the critical study of digital practices related to data and algorithms as well as the examination of implications of data science for privacy and surveillance have attracted considerable attention, not just since recent data scandals that severely harmed public trust in social media and web research.
In the emerging fields of Critical Data Studies (1,2,3) and Critical Algorithm Studies (4,5,6) scholars systematically study and tackle legal, ethical and social challenges of data science. Data scientists themselves have long documented their critical engagement with the creation, collection, storage and analysis of big data and research integrity (7,8,9). However, the focus on important constraints and limitations of computational methods, such as blackboxed methods and opaque access to data, the resulting biases, issues of sampling and representation, external validity and evaluation – just to name a few – is discussed only at the margins of data science. Furthermore, on the solution-oriented, computational end of the spectrum, we often lack the productive intersection of cultures of critique with those of practice.
Therefore, this special issue is dedicated to bringing together critical expertise of scientists in data-driven research areas, who reflect their daily routines, their methods, data sources and the social impact of their research. We would also like to give space to those experiences coming from newly established collaborations of computer scientists with social scientists and humanities’ scholars, moreover with policy makers, activists, or in citizen science projects. The focus is on the critical reflection of scientific methods, data sources, modeling, validation, replication, and review procedures including questions of their impact regarding social behaviour, power relations, ethics, and accountability, thus the performative and normative aspects of data science practices.
The special issue welcomes contributions that utilize data science to engage with methodological, theoretical, practical, and ethical issues of core scientific practices and discuss complementary or alternative routes to robust scientific insights. The special issue seeks to describe and discuss the following topics from a critical data studies perspective:
1. Data collection, modeling, data mining, data sources, and data processing in general
2. Methodological issues, such as bias, limits, scaling, etc. for data mining, machine learning, social network analysis, text analysis
3. Visual analytics and information visualization
4. Reproducibility, archiving, stewardship and digital preservation
5. Data security and protection as well as privacy and data regulation
6. Education and training in critical data science
The editors encourage interdisciplinary co-authorships, e.g. from computer science and critical data studies, as well as reflections on projects on newly developed or prototyped solutions to overcome some of the limitations and issues identified by critical analysis.
Authors can submit abstracts and manuscripts via this website by clicking on "Submit your abstract" and "Submit your manuscript", respectively.
For abstract submission: Title (500 chars max) and text (1000 words max).
For manuscript submission, please follow the author guidelines: https://www.frontiersin.org/about/author-guidelines.
Progressing digitalization, pervasive mobile computing, and omnipresent social media platforms with billions of users have led to the rise of big data analytics and the implementation of data-driven research approaches in a wide variety of scientific fields. Recently, the critical study of digital practices related to data and algorithms as well as the examination of implications of data science for privacy and surveillance have attracted considerable attention, not just since recent data scandals that severely harmed public trust in social media and web research.
In the emerging fields of Critical Data Studies (1,2,3) and Critical Algorithm Studies (4,5,6) scholars systematically study and tackle legal, ethical and social challenges of data science. Data scientists themselves have long documented their critical engagement with the creation, collection, storage and analysis of big data and research integrity (7,8,9). However, the focus on important constraints and limitations of computational methods, such as blackboxed methods and opaque access to data, the resulting biases, issues of sampling and representation, external validity and evaluation – just to name a few – is discussed only at the margins of data science. Furthermore, on the solution-oriented, computational end of the spectrum, we often lack the productive intersection of cultures of critique with those of practice.
Therefore, this special issue is dedicated to bringing together critical expertise of scientists in data-driven research areas, who reflect their daily routines, their methods, data sources and the social impact of their research. We would also like to give space to those experiences coming from newly established collaborations of computer scientists with social scientists and humanities’ scholars, moreover with policy makers, activists, or in citizen science projects. The focus is on the critical reflection of scientific methods, data sources, modeling, validation, replication, and review procedures including questions of their impact regarding social behaviour, power relations, ethics, and accountability, thus the performative and normative aspects of data science practices.
The special issue welcomes contributions that utilize data science to engage with methodological, theoretical, practical, and ethical issues of core scientific practices and discuss complementary or alternative routes to robust scientific insights. The special issue seeks to describe and discuss the following topics from a critical data studies perspective:
1. Data collection, modeling, data mining, data sources, and data processing in general
2. Methodological issues, such as bias, limits, scaling, etc. for data mining, machine learning, social network analysis, text analysis
3. Visual analytics and information visualization
4. Reproducibility, archiving, stewardship and digital preservation
5. Data security and protection as well as privacy and data regulation
6. Education and training in critical data science
The editors encourage interdisciplinary co-authorships, e.g. from computer science and critical data studies, as well as reflections on projects on newly developed or prototyped solutions to overcome some of the limitations and issues identified by critical analysis.
Authors can submit abstracts and manuscripts via this website by clicking on "Submit your abstract" and "Submit your manuscript", respectively.
For abstract submission: Title (500 chars max) and text (1000 words max).
For manuscript submission, please follow the author guidelines: https://www.frontiersin.org/about/author-guidelines.
![By varying the precision of the predicted group non-A one can vary the size of the group A and retain the same precision for group A. Orange dots are in group A (30%), blue dots are not (70%). The two predictions (n) and (o) both have precision of 80 % for group A, yet different accuracy [88% for (n) and 76% for (o)], and errors.](https://www.frontiersin.org/_rtmag/_next/image?url=https%3A%2F%2Fwww.frontiersin.org%2Ffiles%2FArticles%2F502780%2Ffdata-03-00005-HTML%2Fimage_m%2Ffdata-03-00005-g001.jpg&w=3840&q=75)
