|
|
|
|
New Project Helps Teach Undergraduates by Susan Kalish A new project, based at the University of Michigan Population StudiesCenter,will provide the materials to give college freshmen and sophomoreshands-on experience in analyzing U.S. Census Bureau data. Sociolog y professorWilliam H. Frey has been awarded a three-year grant from the U.S. Departmentof Education Fund for the Improvement of Post Secondary Education (FIPSE)to package and disseminate to campuses around the country an educationalapproach that he has be en using since 1988. "This is an unusual approach. In most universities, students don't get their feet wet in data until they learn statistical skills, which generally wouldn't happen before the junior or senior year," says Frey. "Specialized data analysis skills are generally taught in upper-levelmethods courses, but detached from the content questions that hook student interest," he adds. "The aim of this project is to get students working with data early, and in the context of questions they are interested in, by adding a data analysis component to courses on the family, or women's roles, or African-American studies, or other topics." 
TheFIPSE-supported project is developing data modules and a workbook of exercises to be shared with a network of interested faculty around the country. The topics of the first two modules, to be completed by the start of the 1995-1996 academic year, are racial/ethnic inequality and gender gaps in occupationand earnings. Future modules will focus on the family, the elderly, and cohort demography. Data dome from the U.S. Census from 1950 through 1990. Under the project, data will be made available through the Internet. An on-line conference with a "list server" (automated recipientlist) will create what Frey calls an ongoing "virtual workshop"- providing a venue for participating faculty to receive updated materials, share experiences, and work out their problems. There are no charges to participate in the program or receive materials. However, instructors are required to report on their use of the prototype materials. Also, students must purchase the commercial data analysis software package at a cost of about $5.00 per student. Usable High Tech The FIPSE-supported project builds on Frey's seven years of experience in teaching University of Michigan freshmen and sophomores the practical and conceptual skills that they need to use demographic data to investigatesocial issues. Students work in teams of three. The peer support seemsto help students get comfortable with the new computer skills and with posing their own questions to a set of data. Students interact with the data through an extremely user-friendly computer program called CHIPendale. The program enables novices to run cross-tabulations of census data and present these in table or chart format. Through skill-building exercises, students practice controlling for different variables and observing the effect. In fact, without acquiring this basic sense of the logic of data, even advanced students wielding sophisticated statistical techniques often go wrong. The course provides training in the use of technology that is now readily accessibled - but often under utilized - in university computer labs. "Undergraduates tend to use computers for word processing or recreation," comments Frey, "not as tools for analysis." Ready to Think About Data Most freshmen and sophomores may not be ready for a high-powered statistics package like SAS or SPSS (the standard statistical packages for social sciences research). They are ready, however, to observe distribution patterns in a percentage table or a bar chart, to pose a sociological questions, and use data to test a hypothesis. They are ready to learn how to construct proofs that rely on presenting data in a logical way. Another benefit of such a course is that students become literate in common quantitative measures - such as birth rate, poverty, or household. In general, students come to appreciate data as a tool for understanding how a variety of societal changes are adopted and transmitted across different groups in the population. "What is really being taught is the logic of social science inquiry," comments Frey. "That is, first to look for differences that we want to explain. Then, to control the data in some way - along some variable. If the differences go away, your hypothesis is right. If the differences do not disappear or are not appreciably reduced, then you have to look for another explanation to test." In addition to the U.S. Department of Education/FIPSE program, other funders include the University of Michigan and the Alfred P. Sloan Foundation. The Great Lakes Colleges Association helped develop and pretest the modules. For more information, contact Will iam H. Frey, Population Studies Center, University of Michigan, e-mail: bill.frey@usa.net. Sidebar: Hands-On Data Analysis This year, University of Michigan students in the Social and Demographic Change in America course - the prototype for the FIPSE modules - saw firsthand how census data could illuminate aspects of the gender gap in occupation and earnings. The data on full-time, year-round employed doctors came from the 5 percent sample of the 1990 Census. It was striking to note the difference in the percentage of women doctors among different age groups: 26 percent of doctors age 25 to 34 are women, compared with 11 percent of doctors age 45 to 54. Since most people complete professional school as young adults, students learn how comparing the characteristics of different age cohorts in one census year can be one way of suggesting changes in occupational attainment over time. The next question to come up was whether the new pattern of occupational attainment erased the gender pay gap. But, a gender wage gap persisted, even among younger physicians age 25 to 34. At the lower end of the wage scale, 61 percent of younger women doctors, but only 48 percent of younger men doctors earned less than $40,000 per year. In the middle range of income, gender representation was much more even, with 32 percent of women and 35 percent of men earning between $40,000 and $99,999. In the upper range, only 7 percent of women compared with 17 percent of men physicians earned $100,000 or more.
For more information about this project, please contact: ssdan-staff@umich.edu. |
|
|
 |
|
 |
||
 | ||||
This article excerpted from Population Today,