Bringing Big Data to I/DD
The term "big data" refers generally to the information generated by the everyday use of digital services, that are produced so frequently, in such diverse ways, and at scales so large that the data cannot be either stored or analyzed in traditional ways. The analysis of this data can provide valuable insights into health and human behavior and is already widely used by business and industry. It is estimated that well over 90 percent of all digital data in the world has been generated in just the past two years.
Big data has also come to science in a big way. This is especially obvious in neuroscience and the genomic sciences, but significant inroads have also occurred in the social and behavioral sciences. The classic cultural metaphor for big data is “Moneyball,” the book and movie about how an early application of big data to the performance of professional baseball players revolutionized the sport.
Big data has also made inroads into the science and treatment of intellectual disabilities. A prime example of this was the invention of LENA, a device that allows the collection and rapid analysis of extraordinary amounts of language and social interaction data between young children and whoever they are interacting with. This has already led to important discoveries and may well lead to more effective early intervention and diagnosis. Because LENA is based on physics of human speech, not a specific language, it is already being used in studies around the world.
Another area of rapid innovation is the collection of real time data on the response of children and adults to teaching and intervention protocols. This data can be collected by handheld devices such as an iPad and uploaded via “The Cloud” to a centralized server. It can then be aggregated in creative ways that can facilitate more precise instruction and the development of more effective data driven education and treatment protocols. Indeed, at the heart of “big data” is the ability to use various digital devices to rapidly collect data, and for this data to be aggregated to create very large data sets that can then be quantitatively analyzed in ways that may lead to more effective diagnoses, education and treatment and ultimately better outcomes for children and families. Today, this is more potential than reality in most cases, but the environment is rapidly changing.
In response to these rapid changes, the National Institutes of Health launched its own “Big Data and Knowledge Initiative” (termed BD2K) with a workshop at NIH held July 29-30, 2013. This event focused on a wide range of challenges that must be overcome by biomedical and behavioral scientists if the Big Data Revolution is to lead to major advances in science and treatment. It was noted by participants that the two main characteristics of big data are that it exceeds the capacity of unaided human cognition for its comprehension and strains the current technology capacity of most labs and clinical programs. The complexity of the data will frequently require new visualization tools to for scientists to make sense of it. Amongst the many other challenges faced by individual research teams will be locating and accessing the data; organizing, managing and processing the data; developing new methods of analyzing the data; and finding and training researchers who can utilize the data effectively. Among the types of big data problems already impacting the world of research on human health and disabilities are the use of electronic health records, imaging data, genomic data, and especially the integration of large and small datasets. Furthermore, successful big data science is a team sport requiring closely-collaborating scientists who have been cross-trained in multiple areas.
Behavioral and social scientists face some of the same challenges with big data that biomedical scientists do. Consequently, the 2015 Gatlinburg Conference will explore a range of issues in the application of big data to the study of children and adults with intellectual and development disabilities. Two of the plenary talks will focus on innovative technologies that have been recently applied to research with children with autism and other disabilities. A third talk will focus on the quantitative and analytic challenges that behavioral scientists must anticipate and address in course of “big data” research.