Correlation Guessing Game

This page authored by Roger Woodard, Steve Stanislav, Jennifer Gratton, Pam Arroway, NC State University, based on an applet by John Marden, University of Illinois Urbana-Champaign.
Author Profile

This material is replicated on a number of sites as part of the SERC Pedagogic Service Project

Summary

In this game activity, students match correlation values with plots generated by the applet. Competition in this game setting encourages students to become more involved in the classroom and attainment of learning objectives. This game is best if used in a lab setting, although it may be modified to fit other classroom situations.





Learning Goals

Students should be able to:
  • distinguish between positive and negative correlations.
  • distinguish between strong and weak associations.
  • match approximate correlation values to particular scatterplots.

Context for Use

This gaming activity is used in a lab setting for an introductory college statistics course. The nature of the lab setting allows direct feedback for an instructor. In addition, if used as a contest among students, the instructor would then be able to verify the longest streak of correct guesses. However, outside of a lab setting, the honor system would have to be used to determine the longest streak of correct guesses. This method creates competition, which in turn motivates the student to achieve the learning objectives. Furthermore, the promise of bonus points (top student, top ten, etc.) will help keep focus on the activity and promote the learning objective. For example, before the activity begins, an instructor will announce that the students who have a run of consecutive matches in the top ten of the class will receive five bonus points.

Prior to this activity students should have some understanding of the relationship between scatterplots and correlation.

Description and Teaching Materials

In a laboratory setting the applet can be a good way to get students involved in the class and learn something too. The instructor should set the stage by reviewing a little about the idea of the correlation coefficient. This review should include explanation of both the sign and magnitude of the correlation. Then the instructor should demonstrate the applet by showing the basic operation of buttons. As a first round it might be useful to have the class as a whole try to decide which correlations match which scatterplots. For example the instructor can take a vote by asking questions such as
"By a show of hands is this a correlation that is positive?....or negative?" "Which of these four plots have the highest correlation?" Next the game can begin. Inform students of the rules of the game:
  • Students will use the applet to match scatterplots and correlations.
  • Students will continue to guess until they miss one.
  • The student with the longest streak will win the grand prize.
  • The top ten students will win the second prize.
  • Once the game begins students may not close and reopen the applet (this will reset the game).
The instructor should then begin the game and monitor the game as it progresses. As with most games many students will get better after the first round and the competition will be more intense. Repeating the game for a second or even a third round can capitalize on that competitive drive. It might also help to increase the level of prizes as the rounds continue.

Materials Needed

  • Computer lab with machines for each student or groups of students.
  • Computer with projector for the instructor.
  • Correlation Guessing Applet

Teaching Notes and Tips

Here are a few tips for using this activity in your classroom:
  • Duration of this activity is flexible. Ideally it should be used in a time setting of 5 to 15 minutes.
  • Although this game may be used for a few rounds, this game is typically not sufficient to fill an entire class period.
  • Use in a computer lab is highly recommended since direct feedback from an instructor is very beneficial. For example, the instructor will be able to point out what exactly is the cause of a minute difference in correlation from one plot to the next should the situation arise.
  • One way to add to the game environment is by giving a prize for the student (or group of students) who achieve the longest run of consecutive matches. If you are feeling particularly generous, have prizes for the top ten students. Typical prize ideas might include candy or bonus points.

Assessment

The assessment of the students learning of this topic can be integrated as part of broader quiz and exam questions that relate to correlations. In those questions students can match scatterplots with the appropriate correlations.
The Example Quiz Questions (Microsoft Word 706kB May3 07) includes examples of integrating the correlation matching with other questions of this type. It also includes a warmup that closely resembles the applet questions and a more challenging version of the game activity that includes eight scatterplots.


Also, students may be asked to plot for themselves, to the best of their ability, a scatterplot given a correlation value.

References and Resources