Full Citation

Title: HIGH-PERFORMANCE COMPUTING AND STATISTICAL ANALYSIS Application of high-performance computing to the management of social science and demographic data

Citation Type: Journal Article

Publication Year: 1997

ISBN:

ISSN:

DOI:

NSFID:

PMCID:

PMID:

Abstract: The costs of conducting censuses and surveys have always been high. Nonetheless, demographic, social , environmental, and health scientists have found it possible to gather quantities of data that far exceed the capabilities of any contemporary computing technology to extract fully the information contained in the data. Recent developments in computing and information science technology offer the promise of supporting innovative approaches to managing and analyzing data that can dramatically increase the ease of access to the information hidden within these data. This presentation describes a highlyintegrated system of data, metadata, instructional tutorials, hardware, and software designed to minimize the technical barriers to working with large and complex data sets. The objective is to provide easy, low-cost, fast, and meaningful access to these data for a broad spectrum of users. 86 Acquiring and managing data, and extracting information from those data, are uniquely human activities. Information is important at every level ofhuman inquiry today, from personal to corporate, local to international. As a consequence, significant commercial as well as academic effort has been directed toward developing computer-based systems that facilitate the organization, storage, retrieval , analysis, and interpretation of information. The project described in this paper is developing an interactive information-retrieval system, PDQ-Explore, optimized to provide fast, easy, low-cost, and meaningful access to data sets ranging up to the size offull national censuses. Extracting information from data is often an iterative process throughout which the parameters for the next cycle are determined from the results ofthe previous cycle. The system presented here has the potential to dramatically impact this process by minimizing the costs and time associated with each iteration. It is an especially valuable tool when used by investigators who have the substantive knowledge needed to make effective use of a large data

Url: https://link.springer.com/content/pdf/10.3758%2FBF03200572.pdf

User Submitted?: No

Authors: Anderson, Albert F

Periodical (Full): Behavior Research Methods, Instruments, & Computers

Issue: 1

Volume: 29

Pages: 86-98

Data Collections: IPUMS CPS

Topics: Population Data Science

Countries: United States