Robin H. Lock
Mathematics Department
St. Lawrence University
Canton, NY 13617

Outline for a talk at the 1996 Joint Statistical Meetings

  • Abstract
  • Types of data sources
  • Experiences with student datasurfing
  • Conclusion

  • ABSTRACT: The explosion of information available through Internet resources such as the World Wide Web provides a host of opportunities for students and faculty to obtain data for projects and classroom examples. Some prime sites for finding good data are explored and pitfalls with obtaining data on-line are discussed along with some experiences with setting students loose on the Web to find their own e-world data.


  • Dataset Archives
  • Pages of Data Links
  • Search Engines
  • Data about the Web


    During the Spring 1996 semester I assigned regression projects to students in each of three different courses. Many students used data from traditional survey and print sources, but 18 of the 54 projects were based on data from the Web.

  • Spring '96 Web data sources
  • Pros & Cons to using Web data


    For students, we have found those already Web-capable often find the process of finding data on the Web to be more convenient and even fun. Even students with no previous Web experience have successfully discovered and used on-line data with a minimum of training. Nevertheless, we are hesitant to require that all students use the Web as a data source. Some topics and temperaments still are more appropriately addressed by traditional print sources, hands-on experimentation, or surveying fellow students on topics of local interest.

    For instructors, the Web provides many avenues for securing real data for classroom use. Sites such as JSE and DASL also contain some useful advice from other statistics instructors. While we have concentrated here on using the Web to find datasets, we must also point out other Internet resources such as

  • Electronic discussion lists related to statistical education (e.g. Edstat-l)
  • On-line journals (e.g. JSE)
  • Availability of instructors via e-mail (e.g.
  • On-line textbooks (e.g. the Electronic Statistics Textbook being developed at UCLA)
  • Access to statistical software providers (e.g. links provided by Stata)
  • On-line demonstrations (e.g. Fitting a Regression Line provided by Gary McClelland)
  • Web-based course materials (e.g. Mosier's MG284 Course at Clarkson University)
  • More links of interest to statistics teachers can be found at my StatLink page.