ABSTRACT
In order to demonstrate the difficult concept of sampling distributions and the logic of hypothesis testing. I begin each term with students taking samples (five samples with n = 5 and five samples with n = 10) from a population of scores. They are told that the scores constitute a population of ‘sweet nothing’ scores (scores on how skilled a particular population is at whispering sweet nothings in their sweetheart’s ear).
They actually work with the samples in various exercises throughout the term, but the most impressive demonstration comes when we speak of sampling distributions. At this point, I have them compute the mean and standard deviation of the means of each sample; this gives them hands-on experience at what we mean by ‘mean of means’ and ‘standard deviation of means’.
In class we discuss which samples have less variance (ie, the samples with n = 10) and why, and what the shape of the distributions of hundreds of samples of n = 5 and n = 10 would look like. Finally, I show them what the population looks like (skew = 0.83) and what the sampling distributions of the mean for 460 samples of n = 5 (skew = 0.485) and n = 10 (skew = 0.331) look like. I have actually received ‘oohs’ and ‘ahhs’ for this demonstration!
