vol. 25 no. 4, December, 2020

Book Reviews

Jemielniak, Dariusz. Thick big data — doing digital social science. Oxford University Press, 2020. xi, 195 pp. ISBN 978-0-19-883971-2. £25.00.

With technological advances, increased online presence, the arrival of social media, and extensive data points captured on users, data on human behaviour, patterns of social relationships, value systems, cultures and more has increased drastically over the past two decades. Such data can be of great value in sociological studies or social sciences in general. Novel ways of engaging with this huge online data invokes ontological and epistemological questions and invites a reconsideration of different schools of thought on studies of social phenomena (e.g., positivism, functionalism, structuralism, etc.) and age-old dichotomies such as macro-micro; objectivity-subjectivity; quantitative-qualitative; and so on. A look back on the history of sociology and the different key figures who have shaped the path (as far back as Comte, to Durkheim, Weber, Mead and others) is fascinating and helps in understanding how different ideas have contributed to the ways in which sociology and related methods have evolved. It was a mini-revelation for me a few years back when a colleague introduced me to Gabriel Tarde and his idea of quantification as interpreted and re-introduced by Latour (Latour, 2010; Latour et al., 2012). Although Tarde’s ideas were not appreciated in his own time, today’s technologies and the available data afford a revival of his thoughts, while providing the means to achieving what he had proposed.

For me duality of structure (e.g., Giddens, 1984) makes better sense than dichotomies of macro/ micro or structure/agency and so on. Inspired by scholars that try to bridge the divide between such dichotomies and Latour’s revival of Tarde’s ideas, I see great value in studies that combine qualitative and quantitative approaches in order to make a better sense of complex social phenomena. It has become increasingly pertinent for sociologists of today to engage with new technologies and ways of conducting research. A problem remains, however. Traditionally, statistical modelling and quantitative expertise have not gone hand in hand with deep knowledge of qualitative approaches. Many social scientists and humanities’ scholars have seen the value of mixed-methods but some find it difficult to navigate the jungle of data sources, tools, approaches, analysis methods and so on, when it comes to big data. It is, therefore, always a pleasure for me to find publications that try to help with such challenges.

The book Thick big data — Doing digital social sciences is a recent find. The author brings together experiences from traditional ethnographic research, quantitative studies and data science in this book which he describes as an “easy, and practical introduction to doing digital social sciences“ (p.v). The book comprises five chapters. The short introductory chapter provides a background or a context for the book. Chapter two, Online revolution, is subdivided into Online relations, The demise of expert knowledge and Sharing economy. Chapter three on Methods of researching online communities is subdivided into Quantitative research, Qualitative research, and Research of works of culture. Chapter four on Research ethics includes Internet as the Source of infamy, Anonymity, Privacy, Informed consent, Data ownership and Data confidentiality. The two-page short fifth chapter is titled Final remarks. The book concludes with a list of References and an Index.

This book did not quite follow the undefined and yet expected structure that floated into my head when reading it. At times I had to pause reading to ponder about who the author may be and which audiences he might have had in mind. The author most certainly possesses valuable experiences and knowledge that he generously shares, however, in terms of form, his choices regarding the structure, contents, or the level of details provided (or not) on each topic were not all that clear to me. Having said that, I found this book very easy to read and highly interesting; the text flew nicely and each page was rich with examples and ideas that kept me interested throughout.

Still, at times, I found myself having small mental conversations with some of the contents of the book. For example, some formulations seemed to indicate an underlying technological determinism (e.g., on page 6 when the author talked about social changes caused by technology) which I would have liked to discuss further. As another example, the use of Chromebook and Google cloud was recommended as a measure towards user data security (on page 128). It was argued that being averse to storing data in the cloud is unjustified. For me, handing over so much data about ourselves, our on-line lives, and now even the research data to a handful of powerful US-based commercial organizations is also a problem worth reflection and discussion. I see dangers in this asymmetrical access to world’s data by just a handful of organization. Furthermore, despite the US being a free democratic country, the user privacy laws and upholding of human rights are not always as advance as one would desire. Therefore, I for one think that there are more discussions to be had and measures to be taken in order for me to feel comfortable in entrusting sensitive information to such services. I also find opacity of black-boxed solutions and algorithms a problem that keeps me away from using some of the solutions out there. I would have welcomed a critical discussion of such issues too in this book.

However, despite a few minor points that invited further discussions, I found multiple strong aspects to this book. Among others, a very useful part of the book is the section on quantitative research where many useful tips on various data sources, as well as tools for data acquisition and analysis are provided. These usually take a long time for a novice to identify and get familiar with. There is enough information on each of these to help the interested reader know what they may offer, to then be followed up in due course if deemed of use. I was familiar with many of these services or resources, but I too found some that whetted my interest and which I will be checking out at some stage. Another strong aspect of the book was in the section on qualitative research where the differences between digital ethnography and its more traditional counterpart are explained well. (In the same section, case studies, online interviews, and narrative analysis are also discussed.) Furthermore, the book is full of interesting research ideas throughout, therefore it would be a good resource for those who may be in need of inspiration. Finally, the book is also very rich in references to related work which will be an asset especially to students who are in search of a good collection of reading material. As a coleader of a social media study research group, I was familiar with many of the works on the list, but still I found many items that I have now noted in my “to read“ list. In fact, with 39 pages of references (as compared to 132 pages for the main body of the book) this list seemed unusually long.

In short, I enjoyed this book and found it useful. I personally will keep it as a reference book to which I will return again for the items that I have marked as resources to check or publications to read or for study ideas that I can discuss in conversations with my students. I therefore, feel very happy to recommend it to a range of different readers who may be interested in doing digital social science and are looking for inspiration and/or valuable insights.


Nasrine Olson

University of Borås
November, 2020

How to cite this review

Olson, N. (2020). Review of: Jemielniak, Dariusz. Thick big data—doing digital social science. Oxford University Press, 2020. Information Research, 25(4), review no. R702 [Retrieved from http://www.informationr.net/ir/reviews/revs702.html]

Information Research is published four times a year by the University of Borås, Allégatan 1, 501 90 Borås, Sweden.