Date/time: 4/14/15 12:30 to 2PM
Developing a New Form of Computational Social Sciences:
Organizational Genetics To Study Digital Innovation
Harry A. Cochran Professor in MIS
Computational Social Science refers to a set of methodologies and theories that attempt to explain human and organizational behaviors using large-scale data sets of observed behaviors. Much of the work in this area has successfully identified important structural elements that define, constrain, and enable how these large-scale behaviors emerge (such as power law distributions, preferential attachment, and degrees of separation in networks). However, we are seeing the emergence of a new class of digitalized products and services that do not adhere to these structural elements, where their evolution is not as yet well-understood, such as: app ecosystems, mash-ups of digital contents, derivative innovations in open source communities. These digital artifacts frequently evolve beyond the original intent of the designers. Such digital artifacts continue to recombine with other digital artifacts, forming complex sociomaterial systems that are constantly evolving. They are characterized by the absence of central governing bodies and coherent design hierarchy that determine the evolutionary path of these artifacts.
As such digital innovations are becoming increasingly frequent, I embarked first on several qualitative studies to understand theoretically their non-linear dynamic nature of their evolutions. From those studies, I, together with colleagues, developed a theory of wakes of digital innovation and a theory of layered modularity. The theory of wakes of digital innovation suggests that digital innovations do not simply get diffused unchanged, as the traditional diffusion of innovation theory suggests. Instead, they continually mutates as they collide and recombine with other digital innovations, forming wakes-like shape of generative diffusion patterns. The theory of layered modularity suggests that in digital ecosystem, systems are not necessarily designed according to a priori design rules that define the modular architecture as the traditional theory of modularity suggests. Instead, complex digital systems emerge through on-going interactions among heterogeneous and independent individual components. Such layered modular systems are characterized by highly generative and unbounded evolutionary pattern.
Quantitatively testing these theories in large datasets has proven to be quite challenging, requiring new collaborations with colleagues in computational biology and graph theoretic computer science to establish the building blocks to study the non-linear evolutionary patterns of digital innovation. This has led to the creation of new form of computational social science that I refer to as organizational genetics (http://orgdna.org and http://playbigdata.org), that is gaining ground in organizational design, IS, and computational biology. In this talk, I briefly explain: the theory of wakes of digital innovation, the layered modularity, and the building blocks that have taken 5 years of NSF-funded support to develop, and early results of some initial theory testing. I will conclude with a recently received NSF grant to expand organizational genetics methods and theory testing with an aim to build a predictive analytical model based on the work that we have done so far. I will conclude the talk with theoretical and practical implications on digital innovations in self-organizing ecosystems.