Home > Chapter 3

**Chapter 3: The Little Figures That Are Not There**

Chapter three addresses the issue of misinterpreting statistics because of hidden figures. Inappropriate sample sizes can lead to this misinterpretation. If a sample size is too small, it won't represent the average or whole as well. Also, when dealing with statistics, often the experiments get rid of the tests that don't show the results they want. Therefore, the data will be skewed because it will not take into account all the different outcomes that happen. This is what happened in the first example in the chapter with the Doakes' toothpaste. In this example only 12 people were tested to collect the data. With such a small sample size, the results will have a huge change if only a few people to have fewer cavities. This company also got rid of the data from the sample groups who didn't show improvement in dental hygiene. In reality however, there was little or no change in the number of cavities after switching to this brand of toothpaste. Another reason the sample size could be too small is if the probability of something to happen were so slim it most likely wouldn't happen in a group that size anyways. When researches use too small of a group in this type of study, they can conclude that a certain drug will prevent a disease. Because it is not probable for the disease to occur in this small of a group anyways though, it's impossible to prove if the drug has any affect.

The term "average" can also lead to misinterpretation and hide figures in statistics. This term can refer to the mean, median, mode, or another type of average. The interpretation of this term is often left up to the one reading about the results, and the average is often calculated a different way that what the reader interprets. For example, the book talked mentioned a study saying that the "average" number of people in a family is 3.6. However, this number was found by calculating the size of most families, which is either three or four. When most people hear "average number of people in a family" they think that it refers to the average number of people in a family so the data gets misinterpreted. One should also not base information off the average because often it is calculated to where half the population falls above that number and half falls below it. Instead, the normal should have a range, not just one number. Another way data can be misinterpreted because of the misuse of numbers is if there are no numbers on the y-axis of a graph so it's impossible to determine how large the change is. With statistics, it is easy to skew the numbers and use them to your advantage by either using the wrong sample size or interpreting "average" in a different way.

Check out these examples: