Statistical Concepts Have Literally Thousands of Applications,

Statistical concepts have literally thousands of applications, but I will focus on those that apply to several major fields: political science, marketing, economics, social services, and insurance. Statistics are so key to the nature of these fields that most of them could not exist without concepts such as the median and sampling.

Political campaigns are designed to appeal to targeted demographics, which form the basis for blocks of voters. Whereas Abraham Lincoln wrote the Gettysburg address on the train to Pennsylvania on the back of an envelope, modern political speeches are designed to specifically appeal to a median group of voters, and to reflect the reasoning skills and personal tastes and values of these voters. A concept like the 'Axis of Evil,' seems adolescent to university professors and political analysts, but speech writers didn't have these people in mind when they created the concept; by definition, the median IQ is 100. Political pundits see their voters in groups like 'Soccer Moms;' concerned mothers who vote in the interest of their families. On any given opinion, these groups constitute modes reflecting a certain perspective.

The underlying context of demographics is perhaps more prominently employed in marketing. When companies that wish to advertise products on television buy time, companies like Nielsen do statistical surveys by sampling television viewers.

This determines the audience's consumption habits so as to maximize revenues. Big budget prime time shows are the most expensive because the most people are more likely to be watching television. Advertisers might show car commercials during this time, because people with steady incomes are more likely to watch television at this time. Conversely, daytime television might be interspersed with ads for psychic hotlines and the DeVry Institute in order to cater to the consumption habits of the unemployed or those that are employed part time.

When marketers introduce a new product, they often put it in front of focus groups that represent a sample of the general population. For instance, when a telephone company introduces a touch-tone menu for customer service, it will put several different potential menus in front of focus groups. If a system isn't intuitive to most people, they scrap it and use another system. This methodology is used across all new products, from movies to orange juice to commercials and advertisements. It is the inspiration for annoying telephone surveys.

When a marketer approaches someone and offers them a token amount of money to sample a new cola product, her invite is often accompanied by questions such as 'are you the primary wage-earner in your household?' And 'how old are you' and 'how many times a week do you consume carbonated beverages?' These questions are asked for sampling reasons: companies know that if only 20-something diet-conscious women will buy the granola bar they are selling, that it makes sense to advertise in markets where these women are statistically likely to live. Hypothesis testing might include the trial release of such a product in several cities before the product is 'rolled out' nationally. Often cities will be statistically similar to a hypothetical 'average' American city: marketers know that if a cities consuming habits deviate too far from the middle that it is more risky to over-produce the product.

Economics has a strong statistical component. When investors look to see whether or not the economy is doing well, they often pay attention to 'leading economic indicators' such as indexes that reflect consumer confidence. These indexes are based on statistical sampling. When investors purchase debt in the form of bonds, these bonds have ratings such as A+ and B- that represent how risky they are. These ratings are given debt by independent ratings agencies: Standard & Poor's, Fitch, and Moody's. They represent the statistical probability of the company not being able to pay back the price of the bond, which is called a default. For instance, Municipal Bonds are issued by Municipal Governments; only two municipalities have ever defaulted on their bonds. Investors are only willing to invest in 'speculative' or junk bonds if they offer a superior return. Hypothesis testing that is done to determine the risk of a default usually includes revenue projections. These consist of a statistical mean - a 'base case' scenario, as well as a worst case and a best case scenario.

The delivery of suitable social services is also predicated on statistical data. This data is collected once every ten years during the decennial census. All Americans get a 'basic' census form, but some are also given a 'long' form. The U.S. Government then imputes different information from this data, which affects things such as school and library funding, which is done on a county-by-county basis.