 |
|
Editor: Joanne Juhnke, Special Librarian December 2004
Internet Corner (Post Election Roundup) BEFORE: Electoral Vote Predictor How do people use their time? Many surveys ask questions about how much time certain groups spend doing specific activities. Education surveys ask how much time students spend doing homework. Media surveys track TV-watching time and Internet use time. Even the 2000 Census had a time-use question: how many minutes in the commute to work? To study the broad question of how people use their time, though, it takes a time-use survey. A time-use survey specifies a block of time, often a 24-hour day, and asks respondents to recount what they did during those hours and how long each activity took. The University of Michigan and the University of Maryland have conducted national time-use surveys on occasion in the past three decades, addressing the question of how Americans use their time. This year the federal government has taken on the task of collecting time-use data on a nationwide scale. In September 2004, the U.S. Bureau of Labor Statistics released the first annual estimates from the American Time Use Survey (ATUS). The published tables include major time-use categories and some more detailed activities, and selected demographic and labor force characteristics. Microdata files for public use will be released online later. The ATUS sample comes from households that have completed their final month of interviews from the Current Population Survey (CPS), also from the Bureau of Labor Statistics. Data collection, carried out by the U.S. Census Bureau, started in January 2003, with over a thousand interviews completed each month. Like the CPS, the ATUS will be a continuous survey. Time-series comparisons will be possible as more years of data are collected. Why does the Bureau of Labor Statistics want to gather information on what goes into every hour of the average American’s day? A crucial part of the answer is: unpaid work. “Non-market” work activities such as housework, volunteer work, and child care currently do not figure into estimates of economic output such as the GDP. Collecting data on unpaid work may set the stage for revised or additional output estimates. Researchers will have plenty of angles other than production estimates to examine, however. The responses describe not only how people spend their time, but with whom they spend it and where they do their work. In addition, nearly fifty other countries have conducted or have plans for time-use surveys. The ATUS was designed with cross-country comparisons in mind. The American Time Use Survey information and results are online at http://www.bls.gov/tus/home.htm. Allied Drive MSCR Outreach Event On November 22nd DPLS and members of the Data & Computation Center presented a story hour program to thirty-five K-2 grade school children attending the Allied Drive Safe Haven/ Madison Schools Community Recreation (MSCR) program. A collaborative effort of the UW Special Purpose Libraries, the story hour pilot project was implemented by the Water Resources Library in the summer of 2004 to help bridge the achievement gap in literacy experienced by children of the neighborhood. The story hour program will continue monthly throughout the 2004/2005 year, with each session hosted by a different special purpose library.
Since our primary users are graduate students, the challenge for DPLS was to come up with material that would appeal to the “smallest” of users. The result was a 90 minute presentation based ever so loosely on numbers, with a generous sprinkling of rhyme, animals, laughter, and mayhem.
Starting with the interactive games Going on a Bear Hunt and Rainstorm led by Special Librarian Joanne Juhnke, we segued to I Know Two Who Said Moo, So Many Bunnies, and The Gingerbread Man read aloud by Department Administrator Jean Mindel and library school student Lynn Schneider, with hand puppets activated by Senior Special Librarian Cindy Severt.
An arts project in which each child decorated a paper doll figure was sweetened with homemade gingerbread cookies and juice, and ably assisted by Senior Special Librarian Lu Chou, System Administrator Brian De Smet, and Statistical Consultant Doug Hemken.
The presentation wrapped up with each child “populating” a poster board rendering of a community by placing his or her paper doll in a location of choice: school, playground, swimming pool…even a library! Residential Finance Survey, 2001 The 2001 Residential Finance Survey (RFS) was sponsored by the Department of Housing and Urban Development (HUD) and conducted by the Census Bureau. The RFS is a follow-up survey to the 2000 decennial census designed to collect, process, and produce information about the financing of all non-farm, residential properties. Its purpose is to measure the levels of residential mortgage debt and to provide data for assessing the effectiveness of the current residential finance system in promoting the goal of a decent home and suitable living environment for every American. DPLS staff has downloaded this dataset from the Census web site, http://www.census.gov/hhes/www/rfs/rfs.html and stored it on a CD-ROM for our users. Researcher's Notes I am a second year graduate student in the economics program and am currently working on empirical papers about Big Macs and child obesity. While most people expect the topics to be related, I’m examining two different questions. My first paper tests purchasing power parity. Purchasing power parity (PPP) is one of the foundations of international economics. It’s the idea that the price of the same product in two different countries should be identical once converted into a common currency. Economists generally test PPP by using a broad market basket of goods like the consumer price index. Unfortunately, consumer price indexes aren’t constructed in the same way for each country, so studies using this measure don’t really compare the same basket across countries. This is where the Big Mac comes in. McDonald’s makes the Big Mac in a consistent way all over the world, so it’s a good candidate for testing PPP. I am working with a dataset of Big Mac prices from over 40 countries to see if PPP holds. My second paper is about forecasting child obesity. I am using the National Longitudinal Survey of Youth (NLSY) to put together a model that predicts child and adolescent obesity. Since this is my first experience with the NLSY, I have made many trips to the DPLS asking questions about the survey and reading through the collection of books about it.
DPLS has recently received a single-user subscription to China Data Online from the China Data Center at the University of Michigan in Ann Arbor. The subscription covers access to: annual macro-economy statistics (by province, since 1949); monthly macro-economy statistics (the latest 12 months); county statistics (since 1997); city statistics (since 1996); industrial statistics (monthly since Nov. 2001); and monthly text reports on economic development. The China Data Online site is located at http://chinadataonline.org/. To gain access to the subscription data, please come to DPLS to be logged in at one of our public access work stations. In addition, the subscription covers only the “Databases” section of the site, and not the census or statistical yearbooks data. Internet Corner: Post-Election Roundup In the wake of the U.S. November 2004 election, the Internet Corner takes a look back at some sites for election and poll results. BEFORE: The Electoral Vote Predictor The Electoral Vote Predictor site, at http://www.electoral-vote.com/index.html, compiled state-by-state pre-election polls from multiple sources and parlayed the results into vote predictions for the Electoral College. Now that the election is over, the site serves as an interesting post-mortem on the poll numbers. All of the polling results are available as .xls and .csv files. Other interesting features include a comparison of results by pollster, and an animated map depicting the poll result predictions through time. The Electoral Vote Predictor is maintained by Andrew Tanenbaum, an American expatriate living in the Netherlands, who revealed his identity and political leanings (Democrat) only days before the election. The poll-related content of the site is non-partisan, but don’t expect the same from the links to political humor and political sites. In 2000, television networks used exit poll data to project election winners, only to have to retract their projections as the real votes were tallied. In 2002, the computers failed and the networks had to operate without exit poll data on election night. Now in 2004, exit polling has once again provided an extra dose of controversy. This year the networks formed a new alliance called the National Election Pool (NEP) to replace the Voter News Service that had melted down so spectacularly in the previous elections. The National Election Pool hired Edison Media Research and Mitofsky International to provide exit polling (http://www.edisonresearch.com/politics_campaign.html). The 2004 controversy revolves around the early exit poll results intended for media use, which “leaked” and appeared on various web sites on election night. The leaked preliminary exit poll data differed from the final exit poll data and actual vote counts, with early polls showing John Kerry in the lead though the eventual vote went for George W. Bush. Some have used the discrepancies to call for recounts, while others are using the evolving results as an object lesson in margin of error and the danger of placing too much emphasis on preliminary figures. The final exit poll tables may be found on the MSNBC web site, at http://www.msnbc.msn.com/id/5297118/. The Presidential exit poll results are available nationally and by state, with statewide results for Senate and Governor’s races as well. AFTER: Dave Leip's U.S. Election Atlas The U.S. Election Atlas provides some election data free and some for a fee. The free offerings include an electoral vote calculator for 2004, as well as presidential election results nationally and by state through the course of US history, and by county for some years. However, the data is not in an easily-manipulable form and breaks out only the major candidates, particularly at the state level, making it more useful for look-ups than for analysis. Excel-format data files are available for purchase. Be forewarned: the U.S. Election Atlas goes against the prevailing “red state, blue state” convention, coloring its maps blue for Republicans and red for Democrats. Dave Leip’s U.S. Election Atlas is available online at http://www.uselectionatlas.org/. |
©2008 Data and
Information Services Center, University of Wisconsin-Madison.
If you have trouble accessing this page, please contact disc@mailplus.wisc.edu.