Type:
Category:
Volume:
34(1)
Pages:
22-23
Year:
1980
Publisher:
The American Statistician
Abstract:
It is argued that a nonparametric framework for the introductory statistics course is both mathematically and conceptually simpler than the customary normal-theory framework. Two examples are considered: the Kendall rank correlation coefficient versus the Pearson product-moment correlation coefficient, and the confidence interval for the median versus the confidence interval based on the one sample t statistic.
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education