The PERDI Game - A Detailed Description

The goal of the game is to answer questions dealing with these topics (Probability, Experimental Design, Regression, Descriptive Stats, and Inference) and, like BINGO, get a line of correct answers either horizontally, vertically, or diagonally.

Here is a typical PERDI Card:

 P E R D I 2 10 13 15 23 4 8 12 18 24 3 6 Free 19 22 1 7 14 16 20 5 9 11 17 21

Each student receives a different card. There are card generator programs in the free downloads. Numbers are chosen at random and the corresponding problem will be projected on your computer screen. The type is large and navigating between problems is easy. Problems are multiple choice (A, B, C, D, E) and constructed to be answered within several minutes.

Answers are then discussed and if a student gets the question correct, he circles the corresponding square on the PERDI card. Whoever gets 5 in a row, horizontally, vertically, or diagonally, or any other pattern you decide wins. The game can be played with individual students playing or students grouped together in teams.

There are four games of PERDI available on this website and best of all, they are free. Version four is somewhat more challenging than the other versions. You and your students are free to determine the answers yourself or purchase them at a nominal cost.

Here is a sample Inference question:

 I – 24   A company claims that its brand of ink-jet printer can print out 75 four by 6 photos before it needs a new ink cartridge. A consumer’s research company believes this is too high and does a study. It uses 20 of the same modem printer and finds that the average number of photos that can be printed with one ink cartridge is 73.4 with a standard deviation of 3.3. What would be the conclusion of the research if they wish to be 95% accurate?   A) The printer company’s claim should be rejected because the sample mean of 73.4 is less than the claimed value of 75 and using 20 printers is sufficient in the test. B) The printer company’s claim should not be rejected because the p-value of .022 is too small. C) There is not sufficient evidence to reject the printer company’s claim. 73.4 is too close to the claimed value of 75. D) The t-value of -2.168 with degrees of freedom 19 is sufficient evidence to reject the printer company’s claim. E) The p-value of .022 with degrees of freedom 19 is sufficient evidence to prove that the printer company claim is false.   Solution:   (D) This is a 1-sample t-test. Here are the results:     A) is close to being correct but the t-test takes into account the sample standard deviation s, and not just n. With a p-value of .022, we can reject the null hypothesis that the company’s claim of 75 pictures per cartridge and conclude that the average is less (ruling out B and C). Note that E contains a trap: the word “prove” should be a red flag here. So the t-value of -2.168 (which would then lead to the p-value of .022) is the correct answer here.