One grand aspect of the Internet is the availability of huge data sets that can be employed to understand our world. They have predictive power as well as descriptive value and can serve social studies, science and maths quite well. We hope to show students how to explore relationships between variables and consider causation.

To simulate this opportunity with a CD and no live Internet connection, you will work with a sample downloaded from the Net.

Imagine your team has been hired by the Minister of Health of an African Nation to recommend spending on programs that might bring down the infant mortality rate for that country.

You will calculate the correlation between each of the variables in a database about African nations to help guide your thinking and recommendations. Which variables seem most worthy of governmental action and support?

Look here to compare the countries by country names.

Look here to compare the countries by factors.

1. Download the file "unicef.xls" from the "numbers" directory. Right mouse click (Windows) or CTRL click (Mac).
2. Calculate the correlation for each of the variables in the data set.
3. Determine which variables have the strongest correlation.
4. Which of these relationships might indicate causation?
5. Which of these relationships may show association without causation?

Source of data: Unicef at http://www.unicef.org/statistics/index_24183.html

© Jamie McKenzie, 2005, all rights reserved. No copies can be made or distributed in any format without the express written permission of the author. This CD is intended for use by participants during a workshop in Perth on May 27 only.