1. Compute the Chi Square test using a hand calculator and see if your result matches the applet's result.
2. The intervals for the normal distribution frequencies are:
| lower limit | Upper Limit |
| none | -2.0 |
| -2.0 | -1.5 |
| -1.5 | -1.0 |
| -1.0 | -0.5 |
| -0.5 | 0.0 |
| 0.0 | 0.5 |
| 0.5 | 1.0 |
| 1.0 | 1.5 |
| 1.5 | 2.0 |
| 2.0 | none |
Note that since a normal distribution is continuous, no scores will fall exactly on the interval boundaries.
Compute the expected frequencies based on the normal distribution and see if they match those shown in the applet.
3. Set the actual distribution to "uniform" and keep sampling until you find a sample with a probability value of about 0.50 (for deviations from a uniform distribution). Record these results and show them to a friend who hasn't taken statistics (Don't show the statistical analysis). Ask the friend how often they would expect to find such large deviations from a uniform distribution if the numbers were really sampled from a uniform distribution.
4. What are the assumption(s) of the Chi Square Test?
5. In the Texas lottery, balls with numbers ranging from 1-50 are put in a bin and 6 are drawn without replacement. To see if some numbers came up more frequently than expected by chance, you could take all the numbers selected in the last year and count the frequencies of each of the 50 possible outcomes. Would any assumption of the Chi Square test be violated if you used it to test for deviations from a uniform distribution? If so, would the violation tend to increase or decrease the value of Chi Square?
6. There were 104 drawings in the Texas lottery in 1998. For each of the first 100 drawings, select one of the 6 numbers at random and count the number that fall in each of the following intervals: 1-5, 6-10, 11-15, 16-20, 21-25, 26-30, 31-35, 36-40, 41-45, 46-50. Test whether the numbers deviate significantly from a uniform distribution (for which the expected frequency for each interval is 10 since there are 100 numbers and 10 intervals).
7. Compute the mean of the six numbers drawn for each of the first 100 drawings of 1998. Test whether the numbers deviate from a normal distribution. Hint: Standardize the 100 means and use the expected values shown in the simulation with the intervals shown in Question 2 on this page.