# WWW RESOURCES FOR TEACHING STATISTICS

Robin H. Lock
Mathematics Department
St. Lawrence University
Canton, NY 13617 USA
rlock@vm.stlawu.edu

OUTLINE FOR THE TALK
Presented at
TECHNOLOGY IN STATISTICS EDUCATION
A One-Day Conference for Teachers of Statistics
Boston Chapter of the American Statistical Association
Saturday, March 28, 1998 at Babson College

1. Introduction
2. On-line Course Materials
3. On-line Texts
4. JAVA Demonstrations
5. Electronic Journals
6. Electronic Discussion Lists
7. Data, Data, and More Data
7.1 Dataset Archives
7.2 Pages of Links to Datasites
7.3 Government and Official Agencies
7.4 Data About the Web
7.5 Textbook data

9. Conclusion

### Section 2- On-line Course Materials

More and more instructors are providing course materials to their students through websites. In addition to providing convenient access for students, these pages can be perused by instructors at other institutions looking for hints and ideas to improving their own courses.

General Introductory

AP Statistics (High School)
Linear Models
Multivariate
Psychology
Education
Lots of Courses

### Section 3- On-line Texts

Several individuals and groups have undertaken ambitious projects to develop statistics textbooks that can be accessed via the Web. Some examples...

### Section 4- JAVA Demonstrations

The emergence of JAVA as a platform-independent Web programming language has encouraged individuals to develop interactive demonstration software which can be accessed over the Web. NOTE: Although you can link to these sites with almost any Web browser, you must use a JAVA-capable browser to see and execute the applets.

Some of our favorite individual JAVA applets (selected from the sites listed below):

• Guessing Correlations - a neat "game" to show the relationship between correlations and scatterplots - ten minutes should get the ideas across to almost any student. Part of the CUWU Statistical Program at Illinois-Champaign-Urbana.
• Regression - see the effects of adding an outlier (Webster West - S. Carolina).
• Histogram - check the effect of bin size on Old Faithful data (Webster West - S. Carolina).
• Effect of a "Small" Effect - visualization of the impact of a difference in group means on job candidate selection (Mark Lane - Rice).
• Power of a Hypothesis Test - nice graphical look at power for a test of a mean (Todd Ogden - S. Carolina).
• Monty Hall Problem - as implemented at UIUC lets you see how everyone else has done, while the
• Let's Make a Deal version done by Webster West is a bit quicker to play and includes an explanation.
The sites below contain collections of  JAVA applets or links to applets at various locations which are relevant to statistics instruction.
Statistics Packages available as JAVA applications over the Web:
• Statlets - a multi-featured JAVA-based statistics package, with free WWW access for small datasets (10 variables x 100 cases). Produced by NWP Associates, Inc.
• WebStat - created by Webster West at the University of South Carolina.
• Statiscope - a one page univariate stat pakage done by Mikael Bonnier - Lund, Sweden.
Non-java demonstrations: Java is not the only method for showing "live" demonstrations

### Section 6 - Electronic Discussion Lists

These e-lists allow instructors to share questions, ideas and announcements related to teaching statistics, practicing statistics, and statistical computing.  Although e-mail discussion lists are part of the wider Internet resources, the websites which archive the messages also provide good resources for searching through past discussions.
Archives of Apstat-l   This list was set-up to support high school teachers doing an Advanced Placement course in Statistics.  Lots of good conversation about teaching an introductory course.

Archives of Edstat-l (and several other discussion lists) are maintained by the JSE Information Service. This was one of the first electronic discussion lists devoted to teaching issues.  To subscribe send the message

subscribe edstat-l Your Name
to listserv@jse.stat.ncsu.edu. You may also participate through UseNet News at sci.stat.edu..

Mike Fuller's List of Statistical Lists  A number of other e-mail discussion lists are maintained for specific countries (e.g. Australia, Brazil, Italy, New Zealand), statistical topics (e.g. Bayesian methods, experimental design, time series analysis), subject areas (e.g. biometry, geography, econometrics), and software packages (e.g. Minitab, SPSS, SAS).

### Section 7- Data, Data, and More Data

Need an example of a regression which is drastically altered by an influential outlier? Want to find some data to illustrate descriptive statistics which will appeal to students interested in environmental issues? Looking for a multivariate dataset to serve as the basis for next week's midterm exam? Have a student who loves horse racing and wants to use data from past runnings of the Kentucky Derby for her project? The Web is the place to find loads of data sources, often in downloadable formats.

We'll divide our look at Web data into different types of resources:

7.1 Dataset Archives
7.2 Pages of Links to Datasites
7.3 Government and Official Agencies
7.4 Data About the Web
7.5 Textbook data

### Section 7.3- Government & Official Agencies Data

There's tons of data produced by various govenment departments, although it often takes a bit of digging to find it.
The World Factbook - produced by the CIA.  Lots of statistics for each country in the world.  Similar information could be found in a good almanac, but students seem to like browsing for it on the Web.

Bureau of Labor Statistics - lots of economic data -- an especially good source for time series data.

State Data Centers - a list maintained at the Census Bureau to give links to various official data sources in each state.

International Statistical Agencies - a similar Census Bureau list with links to official statistics branches in most countries.

### Section 7.4- Data About the Web

Want to measure some feature of Web pages? You'll need a way to randomly sample pages...
How do colleges and universities present themselves on the Web? Sample from the list at

### Section 7.5- Textbook Data

Data from a number of popular textbooks can now be found on the Web. Note that in most cases, you'll need a copy of the book in order to place the proper context with the data.

### Section 8- Miscellaneous Links

Sites with good general links related to teaching statistics:
Need some questions for that next quiz or exam, check:
To find a statistical software provider's webpage:
Information on statistics textbooks (including reviews):
Other sites to start browsing

### Section 9- Conclusion

The development of the World Wide Web has produced unprecedented global means for instructors to easily share their ideas on ways to improve the teaching of statistics. Although the volume of on-line material may seem daunting, and the process of searching for worthwhile information can be frustrating, the rewards, both for instructors and our students, can be quite substantial. If current trends continue, universal access to the Web should become easier and more common, on-line applications should become even more sophisticated, and useful resources should continue to appear at a steady rate.

Address for this paper:

# http://it.stlawu.edu/~rlock/tise98

or a onepage version at