The goal of this project is to examine the development and current state of U.S. education research through the lens of computational social science. The dissertation comprises three studies which use distinct data--dissertations, library logs, patents--and methods-- topic models, navigational networks, semantic networks--to approach the field. In the first study, large-scale text analysis is used to identify research trends in 137,024 abstracts from education research dissertations defended between 1980 to 2010. Topics associated with the interpretive approach rose in popularity while the outcomes-oriented paradigm declined. These trends are then connected to students' academic job prospects. Academic employment remained stably associated with topics in the interpretive approach, but their effect is mediated by the prestige of the students' institutions. The relation between topic popularity and employability provides insight into field change, and how the benefits of cultural shifts fall along the lines of institutional power. The second study discusses the procuring of education scholarship in an illegal repository of academic papers. Six months of download logs from the pirate academic library Sci-Hub were used to create a network of journals connected by users' consecutive downloads. This network is examined with Personalized PageRank to identify how education journals connect to important outlets in other fields. The results show that users who download primarily from other subjects agree in using the Journal of Chemical Education, which is a general-purpose source of pedagogical materials for STEM classes in higher education institutions. In addition, the journals' subject-specific rankings do not have a significant association with their h-index. These results suggest that there is a large user base of education scholarship who have a predominant interest in science instruction and very likely lacks institutional access to the resources they need. The last study puts forward a methodology to compute the rate of technological change in ed-tech patents. A dynamic semantic network framework is used to identify themes in a corpus of 21,141 ed-tech patents and determine their relative stability over time. This novel network-based measurement of discourse discontinuity shows that educational technology is becoming less discontinuous. However, at the same time, the number of concepts and the relations between concepts in this ed-tech discourse grow considerably. The results suggest that the decline of drastic discourse discontinuities, which would proxy fast technological change, may be a consequence of the overall increased complexity and breadth of the ed-tech sector. Each study presents a source of complexity that conventionally hinders a reflexive inspection of the history and structure of education research. Then, these studies proceed to address each form of complexity through a combination of unconventional data sources and novel frameworks for pattern recognition. Considered together, the studies demonstrate that complex cultural fields, such as education research, can be represented in a simplified form through computational methods, allowing their structure and growth to be made into an object of deliberation. [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]