The Story of Big Data, With All the Small Details
BIG Data may not be “a piece of cake” but during a presentation by Charles Dukes, students and faculty noted that it can be a slice of pie. On April 11, 2019, Holmes Scholars at Florida Atlantic University (FAU) hosted a seminar, “The story of big data, with all the small details,” featuring Dukes, associate professor in the Department of Exceptional Student Education at UFA. Some 25 doctoral students and faculty, along with a Holmes Scholar from the University of Central Florida, attended the seminar with FAU’s Holmes Coordinator Rangasamy Ramasamy and Holmes Scholars Denise Dowdie, Danna Demezier, Shanett Dean, and Deborah McEwan (pictured above with Dukes). During the seminar, Dukes defined “big data” and explored how such data may be used for social science research. He also shared “big data” links that anyone can access, reviewed primary considerations for its usage, and provided an overview of a current research study with big data.
Defining Big Data
Big data is often defined in terms of the volume, variety, and velocity of data or even by a large “N.” Dukes expanded his discussion of big data in terms of “what” it is and “how” it may be analyzed. When considering two such questions, the focus is on the content of the data set and on what is being done with the data set. Big data is not necessarily quantified by a specific number such as 200 or 500 but can include a value such as 50, depending on the number of variables reviewed in the study. Dukes stressed the importance of using the proper lens to review big data.
Before delving into the specifics of big data, he provided a foundational understanding of two different orientations to social science: 1) positivist social science and 2) interpretivist social science. In general, social science was defined as a scientific study of society and human relationships. With the positivist approach, scientific methods are based on the natural sciences with a falsifiable theory premise. On the other hand, the interpretivist approach may use social critique which is not based on falsifiable theory. Falsifiability assumes that a hypothesis must be supported or refuted with the potential of future evidence, which may falsify a given hypothesis. Dukes subsequently clarified that big data resides under the realm of positivist social science.
Slicing the Pie
One great example that was used during the presentation involved an apple pie. In this example, Dukes demonstrated how an apple pie may be sliced in various ways. The slice of pie represents a unit of measurement and provides a commonality on how to describe a social phenomenon. In addition, the slice of pie places a conceptual fence around the phenomenon for mutual understanding. It provides a way to look at a phenomenon (e.g., self-esteem or resilience) and provides parameters for discussion such as how it may be classified and how it can be generalized (e.g. according to gender, age, place, etc.). This process involves providing conceptual definitions, which describe a phenomenon under study, and operational definitions, which describe the parameters of the phenomenon and how it may be measured.
Shaping the Research
Understanding Big data is one possible method to attempt to understand the social world. Social science tries to explain, predict, and understand. Beginning with important contextual notes around the research process and the role of big data in one’s research, Dukes took participants through a plethora of sources and options for accessing big data for their own projects. These options ranged in level of accessibility as well as price, yet all provided means for students to access the data points that can shape their research for years to come. The presentation ended with a question and answer period with Dukes and his doctoral student, whose research pulled from big data sources.