ABSTRACT

This chapter discusses two main areas of statistics: measuring things, especially measuring relationships between variables; and estimating how likely it is that the measure you have gotten could have occurred by chance. It deals with a single statement that negates all possible versions of the statement using a null hypothesis. For any particular set of assumptions, a sampling distribution shows what proportion of the time each particular result could be expected to occur if the sampling technique chosen were repeated a very large number of times and the set of assumptions were true. In measuring the statistical significance of a relationship, one broad procedure holds for all levels of data, although there are differences in specific techniques. A ubiquitous activity in our society is polling, asking a sample of the population whom they will vote for, whether they support regulation of handguns, or whatever. Autocorrelation occurs when observations are related to some extent but are not identical to each other.