Starburst and Sampling - Cover Sheet

SCI Home | SCI Faculty Page | Case Notes | Reference List | Users Notes

Author - Keegan Greenier/John Shepherd - Spring 2000
This case was developed by Keegan Greenier as he first taught SCI in Fall 1998.  Several other people have used the case since and I (JDS) present my version of it here.  I have a few other people's materials in the user notes.

Case Purpose This exercise introduces statistical sampling and can also be used to introduce confidence intervals and hypothesis testing.

Relation to Texts In Chapter 5, Giere discusses probability theory as a way of introducing sampling and the uncertainty attached to estimates.  Section 5.9 in Giere is a long introduction to sampling in which sample size is varied.  I use this exercise as a replacement for this section of the text.   The main ideas of the text seem to be:

  1. Repeated larger samples yield a smaller range of estimate than do small ones, and
  2. Because of the probabilities associated with sampling, "every estimate is an interval" (page159), so we must interpret differences in estimates in terms of probability (this is the subject of Chapter 6).

Description of Case Materials This case consists coffee cans containing starburst candies that are said to represent students at a university.  The different color candies represent the different classes and males are marked with an X.

Case Use This exercise barely fits in a single 50-minute class, and fits well within a 75-minute class.  Students are given the coffee cans ("Folgers U.", "Maxwell House U.", etc.) and asked to estimate the percentage of the student body that is freshmen with different sample sizes.   The class compares the range of estimates at each sample size to see the effect of sample size.  Then they try to estimate the percentage male students at the university with a sample stratified according to their earlier data.  Lastly they compare their estimates of percentage of males and try to decide which university has the lowest percentage.

Reference List Giere 5.9, Chapter 6