Decisions and Distributions We often gather data to help us make informed decisions. A simple yes or no decision can actually have four results. We can be correct by deciding yes when that is the correct decision or by deciding no when that is the correct decision. On the other hand, we can decide yes when we should have decided no, and we can decide no when we should have decided yes. The wrong decisions are Type I and Type II errors, and statisticians calculate probabilities for each. Most of these probabilities are based on normally distributed data. Data that are not normally distributed will affect the probabilities. The decision maker should recognize when probabilities are being influenced by non-normal distributions. These influences can sometimes be ignored, the data can be transformed to a more normal distribution, or different statistical methods can be used. |