Perceived Documentation Quality of Social Science Data.

Abstract

This study is about user perceived documentation quality of social science data. The goal is to identify impacting factors of perceived documentation quality, and find out how perceived documentation quality affects secondary data use. To help identify the impacting factors of perceived documentation quality, documentation of social science data was investigated as a knowledge transfer channel between data producers and secondary data users. A general knowledge transfer model was formulated based on literature in knowledge management, and then applied to characterize the knowledge senders, receivers, knowledge transferred and knowledge transfer channels in secondary data use. In doing this, four possible impacting factors of perceived documentation quality were identified: data producers’ incentives and ability, data users’ absorptive capacity, the existence of intermediaries between data producers and users, and data’s vulnerability to the tacit knowledge problem. Hypotheses about how each factor affects perceived documentation quality were formulated. Interviews and surveys were conducted with secondary users of social science data. Hypothesis tests proved the effects of the proposed four impacting factors of perceived documentation quality. Data produced for sharing are better documented than data produced for self-use. Users with stronger absorptive capacity tend to perceive the documentation they use as better than users with weaker absorptive capacity. Intermediaries such as data archives have been effective in improving documentation quality of data produced for sharing. Data less vulnerable to tacit knowledge problem, such as quantitative survey data and data about straightforward facts, are perceived as better documented than data more vulnerable to tacit knowledge problem, such as qualitative data. Users’ incentives to use secondary data mostly depend on how well the data fit their information needs rather than documentation quality. A well-documented dataset will not be used if it doesn’t answer users’ research questions. Users will not give up using a dataset simply because it is poorly documented. Users often need to seek information not provided in documentation because of inadequate documentation (insufficient, hard to use, inaccurate), inherent limitations of documentation, users’ low absorptive capacity, convenience and social and psychological reasons. In seeking outside information, users tend to consult multiple sources and channels.