Numeric Data Services and Sources for the General Reference Librarian

Chandos Information Professional Series

Numeric Data Services and Sources for the General Reference Librarian

Lynda M. Kellam

Table of Contents

Cover image

Title page

Copyright

Dedication

List of figures and tables

About the authors

Acknowledgements

Chapter 1: Introduction to data services and sources

Abstract:

History of support for numeric data

Data definitions

Chapter 2: Supporting statistical and numeric data services and sources

Abstract:

Environmental scanning

Levels of service

Models of support

Marketing and assessing data services

Future directions

Chapter 3: Reference and instruction for data sources

Abstract:

The reference interview and data

Data instruction

Statistical and data literacy

Chapter 4: Basic sources for supporting numeric data services

Abstract:

Producers of statistics

Types of sources

Search strategies

List of sources

Quick start to finding statistics and data

International

European Union and United Kingdom

United States

Canada

Other parts of the world

Special topics

Locating spatial data

Chapter 5: Data librarianship: a day in the life

Abstract:

Chapter 6: The future for numeric data services

Abstract:

Visualization

Preservation of data

Data citation

The future is data

Appendix A: Respondents’ institutional profiles and full responses

Appendix B: Selected annotated bibliography

Index

Copyright

Chandos Publishing

Hexagon House

Avenue 4

Station Lane

Witney

Oxford OX28 4BN

UK

Tel: + 44 (0) 1993 848726

Email: info@chandospublishing.com

www.chandospublishing.com

Chandos Publishing is an imprint of Woodhead Publishing Limited

Woodhead Publishing Limited

80 High Street

Sawston

Cambridge CB22 3HJ

UK

Tel: + 44 (0) 1223 499140

Fax: + 44 (0) 1223 832819

www.woodheadpublishing.com

First published in 2011

ISBN:

978 1 84334 580 0

© L.M. Kellam, 2011

British Library Cataloguing-in-Publication Data.

A catalogue record for this book is available from the British Library.

All rights reserved. No part of this publication may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form, or by any means (electronic, mechanical, photocopying, recording or otherwise) without the prior written permission of the Publishers. This publication may not be lent, resold, hired out or otherwise disposed of by way of trade in any form of binding or cover other than that in which it is published without the prior consent of the Publishers. Any person who does any unauthorised act in relation to this publication may be liable to criminal prosecution and civil claims for damages.

The Publishers make no representation, express or implied, with regard to the accuracy of the information contained in this publication and cannot accept any legal responsibility or liability for any errors or omissions.

The material contained in this publication constitutes general guidelines only and does not represent to be advice on any particular matter. No reader or purchaser should act on the basis of material contained in this publication without first taking professional advice appropriate to their particular circumstances. All screenshots in this publication are the copyright of the website owner(s), unless indicated otherwise.

Typeset by RefineCatch Limited, Bungay, Suffolk.

Printed in the UK and USA.

Dedication

Thank you to Daniel, my husband, and my parents for putting up with the process and still loving me. I dedicate this book to Daniel, Becki, and Ed.

List of figures and tables

Figures

1.1. Example of basic input of survey data into a spreadsheet 10

1.2. A bar graph depicting the distribution of student respondents to the survey by class 11

1.3. Fact sheet from American FactFinder showing social characteristics from the 2000 Census for New York, NY 13

6.1. Example of a basic map created using GIS that includes spatial data of the Southeastern United States 163

6.2 and 6.3. Pie charts showing different distributions 165

Tables

1.1. Example of statistics generated from the survey data 11

1.2. Median age by sex for the total population of New York City and for the white population of New York City from the detailed tables in the 2000 Census 14

1.3. Sample code for variable ‘HISPAN’ from the PUMS file for 2000 Census of Population and Housing Technical Documentation (U.S. Census Bureau 2003). These are only a few of the possible categories for ‘HISPAN’ 17

About the authors

Lynda M. Kellam is the Data Services and Government Information Librarian at the University of North Carolina at Greensboro’s University Libraries where she is based in the Reference and Instructional Services Department. In addition to her primary duty as data services librarian, she manages UNCG’s federal and state government depositories, serves as collection and instruction liaison to the Political Science Department and several social sciences programs, and co-ordinates the Reference Internship Program.

She completed her Bachelor of Arts in political science at the University of North Carolina at Greensboro in 1997 and began study in political science at the University of Wisconsin, Madison. Her studies in Madison focused on the countries of the former Yugoslavia and national identity formation. After leaving Wisconsin with a Master’s degree in 2001, she studied in Croatia on a Fulbright student fellowship where she worked with B.a.B.e. (Budi activna, Budi emancipirana), a leading women’s rights group, as a researcher and editor. Working with B.a.B.e. reinforced her passion for assisting others with research and led her to explore the possibility of social science librarianship.

Upon returning to the United States, she enrolled in the Master’s of Library and Information Studies program at the University of North Carolina at Greensboro where she concentrated in academic libraries, social science research, and instruction. After graduation in May 2007 UNCG’s University Libraries hired Lynda as the institution’s first Data Librarian with the charge of crafting the direction of this position. Her focus in the past four years has been on outreach and promotion, collection development, and instruction. As the users at UNCG are relatively diverse (from novice data users to the data savvy), she has had to examine the full range of levels of service.

She has been active in the profession at the state, national, and international levels. She serves on the conference planning committee for the International Association of Social Science Information Services and Technology, the primary data librarianship association, and works closely with the American Library Association’s Government Documents Round Table. She was named an American Library Association Emerging Leader in 2010 and received the Association of College and Research Libraries Librarian Scholarship in 2009. She is also a member of the American Political Science Association.

The author may be contacted at: UNCG, Jackson Library, PO Box 26170, Greensboro NC 27402–6170, USA. E-mail: lmkellam@uncg.edu

Katharin Peter is Social Sciences Data Librarian for the Von KleinSmid Center Library for International and Public Affairs at the University of Southern California where she established library data services in 2007. Prior to USC, Katharin spent 6 years as a research analyst and data professional, conducting research for publication by the National Center for Education Statistics and other agencies. She has a BA in Sociology from the University of California, Santa Cruz and an MLIS from Pratt Institute, New York.

Acknowledgements

‘This was the best IASSIST ever!’ is the final official statement made by the organization’s President at the end of every annual conference. As the leading organization for data professionals and librarians, IASSIST also has a tremendously welcoming group of members. As such, this book would not be possible without the direct help of many and the indirect influence of a few.

I first want to thank Katharin Peter for both contributing to this work with her excellent chapter on data sources, but also for putting up with my unintelligible emails for over a year. I have a tendency of roping her into projects and she always responds with wonderful grace and good humor. Thank you also to the entire group of data librarians mentioned in Appendix A who contributed either by reading chapters or answering the informal interview. Special thanks to Kristin Partlo for her words of wisdom in Chapter 3.

In addition to the contributors there are several data specialists who assisted indirectly with the production of this work. This book is partly the result of my attendance at the ICPSR data librarianship workshop conducted by Chuck Humphrey and Jim Jacobs. They have trained many new data librarians over the years, and without their efforts we would not have a framework for developing data services. With this work, I want a wider audience of non-data librarians to learn from their contributions.

I also want to thank my colleagues at the University of North Carolina at Greensboro for assisting me, especially Susan Farr and Amy Harris. Susan Farr willingly read multiple drafts of this work and deserves most of the credit for its sheer readability. Amy Harris has always provided a willing ear for brainstorming hare-brained ideas. This book is the result of one of those sessions.

1

Introduction to data services and sources

Abstract:

The purpose of this chapter is to provide an introduction to data services within the academic library by describing the history and development of data services in the United States and other countries. The chapter also introduces the primary concepts necessary for understanding numeric data, datasets, and secondary data analysis, including the difference between aggregate statistical products and microdata, the purpose of coding data, and the use of data documentation.

Keywords

history of data services

data services

numeric data

secondary data analysis

description of aggregate statistics

description of microdata

Recent years have seen a proliferation of free and subscription-based numeric data resources on the Web. From the World Bank’s Open Data Initiative to the US Census Bureau’s American FactFinder, facts and figures are readily available on an array of topics. This increase in access is the result of several converging technological advances, which include wider Internet availability and use, better and faster Internet connections, and the ability to create compressed files. All of these changes have decreased download times and increased portability of large amounts of information.

In addition, as it has become easier to access numeric facts and figures, librarians have seen an increasing interest in quantitative information among users with a range of skill sets and knowledge levels. Quantitative information is no longer only the purview of statisticians; more people are interested in the possibility of expressing ideas through numbers. The increasing ease of use of spreadsheet tools like Microsoft Excel, and the explosion of Web 2.0-based data visualization websites such as Many Eyes (http://manyeyes.alphaworks.ibm.com), have allowed more people to see the benefits of using and visually representing data.

With these changes in the access to numeric data, librarians have become central participants in assisting users. Our traditional focus may have been on the written word, but the rise in digital formats and files has carved out a new role for the library, one that supports information in all its forms – from the written word, to the digital image, the streaming media sample, and the numeric data file. Moreover, our promotion of information literacy and emphasis on information-literate users means we need to pay attention to all types of information sources, even the non-textual. Users may not immediately associate numeric data sources with the library, but increasingly, libraries and librarians are being called upon to purchase, support, and archive these sources.

Because libraries have acquired more numeric data sources, patrons have increasing expectations that librarians will provide support for these sources. As faculty and graduate students move from institution to institution, they may have expectations that the services offered at one library will be available at another. The difficulty facing many libraries, especially smaller ones, is that the use of numeric data sources, from the basic statistical database to a large data set, requires basic statistical literacy and potentially more advanced skills. Although some librarians may have had a statistics course as an undergraduate or a research methods course in graduate school (or both), most Library and Information Studies programs do not teach the skills necessary to support numeric data sources. Few new librarians have had the same exposure to data sources and quantitative analysis as they have had to more traditional library principles, such as cataloging standards. Even though users may have increasing expectations about the role of libraries in supporting numeric data sources, our professional education, especially in the United States, has not quite caught up with those expectations.

No matter the size of an institution, librarians can expect an increasing interest in numeric data sources. As user expectations for support increase, the need for advanced skills to support data sources will also rise. In small libraries, the social sciences librarian or the business librarian may serve as a de facto data specialist and assume responsibility for data questions. This model can be problematic if the designated librarian does not have the skills to support numeric data sources and has only limited possibilities for training. Gerhan (1999) notes that librarians without a basic understanding of the use of quantitative analysis will have difficulty asking the appropriate questions when conducting reference interviews.

History of support for numeric data

How did libraries begin to support numeric information? Statistical publications in print or microtext formats have always been a part of library holdings, especially those American libraries involved in the Federal Depository Library Program (http://www.fdlp.gov). From Statistical Abstract of the United States to the print Census volumes, librarians have had experience with these publications for a long time. A shift began with the emergence of Machine Readable Data Files (MRDFs). These MRDFs encoded data about all types of information into a variety of machine-readable formats – from punch cards to tapes to CD-ROMs. As these formats became more accessible and commonplace, the need to support numeric data (in addition to other digital files) increased.

In the 1960s and 1970s many libraries did not have the infrastructure to support numeric data files. Several authors note the lack of infrastructure and expertise in libraries combined with a perceived unwillingness to support these emerging technologies (Heim, 1982; Rowe, 1984; Chiang et al., 1993). This void led some universities, especially large research institutions, to create and house data support centers or archives in academic departments or computing centers, rather than within the libraries. Rowe describes the establishment of the Roper Center for Public Opinion Research and the Inter-university Consortium for Political and Social Research (ICPSR) as ‘the first organized efforts to formalize the distribution of MRDF’ (1984: 327). Created in 1962, the Inter-university Consortium for Political and Social Research housed primarily the ‘American National Election Study’. Over the next few decades its mission grew to the archiving of data from a wide variety of studies and responsibility for instructional support.

In the early 1990s the United States Census Bureau released the 1990 Decennial Census on CD-ROM. As these new formats began arriving in the federal depository units of many libraries, librarians and administrators started to consider support for these data files in addition to the typical statistical sources in print (Geraci et al., 2008; Treadwell and Cogswell, 1994). Similarly, Statistics Canada gave Canadian universities more access to data sources through the Data Liberation Initiative (Geraci et al., 2008). Through the 1990s libraries began creating data centers housed within the library, such as the Machine Readable Data Center at the University of Minnesota, and data retrieval systems, such as INFeRS at Cornell University’s Mann Library (Treadwell and Cogswell, 1994; Chiang et al., 1993). Although data support and data librarian positions began to proliferate, they were generally concentrated in the larger research institutions.¹

With the increase in access to the Internet and the decrease in download times over the past decade, we have seen yet another shift in the approach to data services, at least in the United States. While patrons may have quicker and more immediate access, they may not understand how to use specific sources or have the training necessary to comprehend data. For these reasons, smaller universities and colleges have expressed a need for a librarian with a data background – one who is dedicated to supporting and educating users about numeric information.

In supporting numeric data services, libraries have several possible organizational approaches. They could have a de facto model, with data responsibilities supported by the current social sciences or business librarian. A coalition approach is also possible, in which a library’s reference staff supports basic statistical and some numeric data sources while another unit supports more advanced data tasks. The third model, the creation of a designated data services librarian position, is becoming common at smaller universities and colleges. There are various permutations of this approach – sometimes the entire school has only one data support person, while other institutions may have a designated data librarian in combination with units outside of the library supporting quantitative research. Every approach has strengths and challenges and each institution will need to gauge the most appropriate one for their staff, patrons and environment.

Our goal is to provide general reference librarians with a framework for understanding numeric data services and sources, and to increase the level of awareness of and comfort with this specialization. We will focus primarily on social science data sources throughout this book because of the current shape of data librarianship. Many data services positions have been combined with government information or other social science liaison positions within the library and as such there has been a strong social science bias to their work. The principles and recommendations in this book are certainly applicable across disciplines, and data sources also exist for the hard sciences.² To support these ‘hard science’ sources, a library should consider having both a social science data librarian as well as a