Seeking Equilibrium in Data Reuse: A Study of Knowledge Satisficing

Abstract

Government funding agencies and commissions have proposed that sharing, preserving, and providing access to more scientific research data will lead to increased reuse of data in academic research and result in greater knowledge and new discoveries. However, researchers encounter significant logistical, theoretical, methodological and ethical challenges to reusing data that hinder the achievement of these goals. One of the challenges researchers face is obtaining sufficient knowledge about data and the context of data creation to make a decision to reuse the data in their research. In this dissertation, I report on a mixed methods study to investigate how researchers set limits on the types and amounts of knowledge they obtain about data, and what influences them to do so. A more nuanced understanding of how and why researchers determine such thresholds can inform strategic measures to enhance support for data reuse. My study included a survey and semi-structured interviews and was conducted on a sample of researchers who reused data from the ICPSR data archive. I used Donna Haraway’s theory of situated knowledges and Herbert Simon’s theory of satisficing to develop conceptualizations of data and means of evaluating thresholds of knowledge that researchers obtained about data. I defined a concept called “reuse equilibrium”—when researchers determine data are sufficient to reuse to meet their research goals—and examined whether satisficing was a means by which researchers obtained knowledge to reach reuse equilibrium. I found that researchers lacked knowledge they desired about data and that this lack of knowledge frequently had a negative impact on their research. The type of knowledge researchers most often desired but were unable to obtain was “supplemental” knowledge that was not archived with the data and may never have been collected. While researchers lacked knowledge about the data they desired, I found that satisficing did not accurately represent their behavior in knowledge attainment. Instead, researchers sought to maximize their knowledge of data to meet personal aims (i.e., to reach “personal reuse equilibrium”) in environments characterized by pressures and incentives that favored the achievement of social norms and requirements (i.e., "social reuse equilibrium"). I concluded that an important way to improve the environment for reuse was to assist researchers in obtaining supplemental knowledge about data they desired, thus supporting their achievement of personal equilibrium. This could be done by facilitating more structured and intentional “conversations” between data creators and data reusers with the purpose to influence the data that are created in the first place. My findings about the knowledge researchers lack about data and the ways they seek to obtain it will be of interest to data reusers to gain a broader perspective on their colleagues' experiences. They will also be of interest to data creators, as well as data stewards, publishers, and other data intermediaries, to understand the knowledge researchers desire about data and the role they can play in helping researchers obtain it. Such findings, in addition to those about pressures and considerations in the reuse environment, will be of interest to funders and policy makers to gain insight into the ways current policies, practices, and incentives could be enhanced or changed to maximize the return on investment in primary research.