This applet simulates rolling dice to illustrate the central limit theorem. The user can choose between 1, 2, 6, or 9 dice to roll 1, 5, 20, or 100 times. The distribution is graphically displayed. This applet needs to be resized for optimal viewing.
This activity allows the user to simulate pulling red and green balls out of three boxes. The boxes are pre-arranged so that there are two red balls in one box, two green balls in another, and one green and one red ball in the third. The user can shuffle the order of the boxes and the order of the balls in the boxes. To run in single trial mode, click on one of the box to see if the first ball is green. If it is, click on the box again to see if the second ball is green also. A count will be kept of the results. To run in multiple trial mode, enter the number of trials desired in the box and click on the run multiple trials button. This activity would work well in groups of two to three for about twenty minutes if you use the exploration questions provided and ten minutes otherwise.
This chapter of the HyperStat Online Textbook discusses in detail sampling distributions of various statistics (mean, median, proportions, correlation, etc.), differences between such statistics, the Central Limit Theorem, and standard error, giving formulas, examples, and exercises.
This page will perform an analysis of variance for the situation where there are three independent variables, A, B, and C, each with two levels. The user may enter data directly or copy and paste from a spreadsheet or other application.
In the first simulation, random samples of size n are drawn from the population one sample at a time. With df=3, the critical value of chi-square for significance at or beyond the 0.05 level is 7.815; hence, any calculated value of chi-square equal to or greater than 7.815 is recorded as "significant," while any value smaller than that is noted as "non-significant." The second simulation does the same thing, except that it draws random samples 100 at a time. The Power of the Chi-Square "Goodness of Fit" Test pertains to the questionable common practice of accepting the null hypothesis upon failing to find a significant result in a one- dimensional chi-square test.
The following pages calculate r, r-squared, regression constants, Y residuals, and standard error of estimate for a set of N bivariate values of X and Y, and perform a t-test for the significance of the obtained value of r. Allows for import of raw data from a spreadsheet; for samples of any size, large or small.