ABSTRACT
This chapter is concerned with software for the analysis of survey-type data. We start with the assumption that data have been collected — typically from a questionnaire — in a way sufficiently standardized to allow responses to be recorded using a common coding frame. Usually information will have been sought from a number of different individuals (or other units such as households), often sampled on a random basis, with the aim of understanding how particular outcomes or responses (e.g. amount of income earned, number of hours worked) vary in relation to a variety of other factors. Unless a very limited amount of information has been collected, it will be much quicker and easier to analyze the data using a computer than manually. As well as allowing large amounts of data to be handled with ease, computers also enable data to be edited as it is entered, by including checks for valid values and for consistency in responses.
