Resource Library

This applet shows a scatterplot of height versus foot length. Users can add or delete points and then guess the regression line by clicking "Your Line" and moving the blue regression line. By clicking "Regression Line" users can see the actual regression line. The applet also shows the correlation and R-square for the data as well as the residuals and squared residuals for the guessed regression line and the actual regression line.

This applet allows users to calculate probabilities from a normal distribution. First, set the mean and standard deviation and click "Scale to Fit". Check one of the boxes next to the inequality signs and enter a value for x; the applet will calculate the z-score and cumulative probability (shown in dark blue for top value and pink for the bottom). By clicking both boxes, users can see the probability between two values (in pink) or outside two values (in blue). Click the inequality sign to change the direction of the cumulative probability.

This dataset comes from a study of 23 healthy male subjects in a five period crossover study. Each subject followed a randomly allocated sequence of five treatments. Blood samples were collected and serum ethanol levels were recorded. Questions from this study refer to the drug-alcohol interaction. A text file version of the data is found in the relation link.

This dataset comes from a study of 16 healthy males allocated to 1 of 2 treatment sequences: wafer/tablet or vice versa. Blood samples were taken and plasma levels were recorded. Questions from this study refer to whether the wafer treatment and the tablet treatment are equivalent. A text file version of the data is found in the relation link.

This resource explains margin of error using an example on presidential popularity polls.

This resource discusses sample sizes and how they are chosen.

This site presents may articles on current events and issues that challenge statistics reported in the news. Each article encourages readers to think critically about statistics reported by the media and to look at the whole picture before believing conclusions presented in the news. "Our goals are to correct scientific misinformation in the media resulting from bad science, politics, or a simple lack of information or knowledge."

This site contains links to journals on probability and statistics published around the world. "Some publishers require registration to browse abstracts. Others require a current subscription to the journal by you or your institution. Most browsable titles, abstracts and papers are only for the past year or so."

This collection of datasets, posted by UCLA, is divided into 6 groups: Datasets for Teaching; Data from Books; Data from Consulting Projects; Data from National Statistics Agencies; Social Science Data Archives; Data from US Governmental Agencies. The data from books come from the following authors: Petruccelli, Nandram and Chen; Freedman, Pisani, and Purves; Andrews and Herzberg; Carlson and Thorn; Cox and Snell; Hand, Daly, Lunn, McConway and Ostrowski; and Moore.

This applet allows students to explore three methods for measuring "goodness of fit" of a linear model. Users can manipulate both the data and the regression line to see changes in the square error, the absolute error, and the shortest distance from the data point to the regression line.