Technological Determinism in Social Data Analysis

Abstract

In reviewing the history of social data processing, it is argued that, with the exception of Herman Hollerith's watershed invention of the punched card in the late nineteenth century, most social science computation involved adapting technologies developed for other disciplines to the needs of social science computing. During the 1950s and early 1960s, when social researchers had to rely upon very slow and cumbersome unit record equipment, the analysis had to be very carefully thought out or risk being very inefficient. As general-purpose computers began to become widely diffused, general-purpose statistical packages aided immensely in processing survey research data. Because they were designed to take advantage of computational efficiencies, however, they tended to facilitate, if not encourage, somewhat mindless cross-tabulation strategies of computing "everything by all." It is argued that the nature of the microcomputer may again encourage researchers to be more reflective in designing their data processing.