|
Abstract:
|
One of the most difficult concepts to teach in an introductory statistics course is how to interpret a confidence interval, particularly the concept that a confidence interval may not include the true population mean as determined by the sample. To this end, we propose a fun classroom activity. This activity can be used when teaching Z or T distributions and proportional confidence intervals. At the end of this activity, an illustrative graph is created using confidence intervals calculated by each student. Discussion of this graph will demonstrate the dependency of a confidence interval on its sample; a confidence interval is the probability a given interval contains the mean (and not that the probability that the mean is contained in the interval), the intervals change, the mean does not, and correct calculations do not always produce accurate intervals.
|
