In this exercise, you will generate your own data to investigate whether the mean or the median is the most appropriate measure to use. The MATLAB function randn(m,n)produces anm-by-n matrix of...

In this exercise, you will generate your own data to investigate whether the mean or the median is the most appropriate measure to use. The MATLAB function randn(m,n)produces anm-by-n matrix of pseudo-random numbers with zero mean and standard deviation The numbers are sampled from a normal, or Gaussian, distribution. We will discuss the normal distribution in detail in Section 4.4.1, but for now the important information is that the distribution is symmetric with no outliers. Carry out the following investigation:

Now repeat the above, but after cubing (i.e. raising to the power of 3)the output from randn. Because we are only using a small sample size (30), this will have the effect of introducing a few outliers. Now see if the mean or median is a better measure to use?