Think of these as nice little "games" you can play to learn something about statistics. I didn't write these applets -- I have just found them (or friends of mine found them and told me about them).

Histogram How many intervals you use can affect the shape of the histogram. This really nice applet takes a set of data on the times between eruptions of the "Old Faithful" geyser and lets you adjust the "bin size" for the histogram and immediately see the shape. This is a really nice way to see how much difference the bin size makes. (Use the slider labeled "binwidth".)

Guessing Correlations This is great! They give you four scatterplots and four correlation coefficients. You match them up. Then you can ask for more. The people at the CUWU Statistical Program at Illinois-Champaign-Urbana keep a page of the "top 20" scorers this semester. Maybe you can get onto the list!! This is the best way I know of to get a feel for what correlation coefficient goes with a graph. (The first thing to do is click on "New Graphs".)

Correlations This is pretty neat too. You choose the numerical correlation coefficient and they give you a scatterplot with that coefficient. (Slide your mouse up and down the scale.)

Guessing the Regression Line Here you are given a scatterplot and you can use the mouse to indicate what seems to be the best straight line approximating the data. It will calculate the sum of the squared errors from the line. Then do it again and see if you found a better one. When you want to, you can ask what the minimum sum of the squared errors is, so you can see how close your guesses came to the least squares regression line. This is an excellent way to develop a feel for linear regression! (Click on "Begin" on the left side of the page.)

Outliers and Influential Points in Regression. Use your mouse to add a point to the scatterplot and you see how the least squares regression line changes.

Transformations of variables. Choose from several data sets and several possible transformations of X, Y, or both. Compare scatterplots, correlation coefficients, and regression lines.

Sampling Distributions. Simulate sampling distributions. Use the given population or change it. You may simulate the sampling distribution of the mean, variance, or various other statistics. Compare the sampling distribution that you generate with a normal distribution.

Confidence Interval Illustration. This generates multiple confidence intervals for a given problem, so that you can see what proportion of them cover the population mean. You specify alpha.

Confidence Interval Simulation. Specify the sample size and multiple confidence intervals are computed and graphed at two confidence levels.

Power of a Hypothesis Test. Computing this is a bit beyond the scope of this course. However, you can appreciate the applet, which computes the power for you, even if you don't haven't computed any of these yourself. Investigate and see what it takes to get a test with power that is reasonably high.

Group differences. This applet illustrates the effect of sample size on how large a difference between group means can be noticed.

Chi-Squared Test on 2x2 tables. Specify the population proportions and sample sizes and then a table of simulated data is generated and a chi-square test of independence is performed.

Rice Virtual Lab in Statistics | Robin Lock's summary page

MP's 1342 home | MP's home | ACC MATH 1342 home | ACC Math home | ACC home

Last updated September 10, 2003. Questions, comments, suggestions?