Statistical Analyses and Sampling
Statistical treatments of educational information have two purposes. The first is descriptive--to summarize information in an easily comprehended, quantitative form. The second is inferential--to provide an estimate of how likely a sample of people or events accurately represents a broader population of people or events. The following discussion first identifies a variety of descriptive statistics that are useful in research, then turns to matters of inference. The aim of the chapter is to suggest which types of statistics are most suitable for answering different kinds of questions. However, the aim does not include explaining the mathematical foundations underlying those statistical procedures nor to tracing in detail their methods of computation. Such foundations and computational techniques will be found in the kinds of books listed at the end of the chapter.
The descriptive statistics included in this section are percentages, percentiles, measures of central tendency, measures of variability, measures of skewness, and correlation techniques. The presentation of each type opens with questions that the statistic is designed to answer. The discussion then continues with an explanation of the statistic's nature, its advantages and limitations, and examples of its application in comparative studies.
The research question : What proportion of a variable (such as students, schools, equipment, events, or the like) display a particular characteristic?
Among the 300 female students surveyed, 35% would condemn the war criminal to death. Among the 114 males surveyed, 49% would put him to death.