A credit card bank is investigating the incidence of fraudulent card use. The bank suspects that the type of product bought may provide clues to the fraud. To examine this situation, the bank looks at the Standard Industrial Code (SIC) of the business related to the transaction.
This is a code that was used by the U.S. Census Bureau and Statistics Canada to identify the type of every registered business in North America.2 For example, 1011 designates Meat and Meat Products (except Poultry), 1012 is Poultry Products, 1021 is Fish Products, 1031 is Canned and Preserved Fruits and Vegetables, and 1032 is Frozen Fruits and Vegetables.
A company intern produces the following histogram of the SIC codes for 1536 transactions:
Since 1997 the SIC has been replaced by the NAICS, a code of six letters.
He also reports that the mean SIC is 5823.13 with a standard deviation of 488.17.
a) Comment on any problems you see with the use of the mean and standard deviation as summary statistics.
b) How well do you think the Normal model will work on these data? Explain.