This book began with initial contacts between linguists (Hughes and Fletcher) and a statistician (Woods) over specific research problems in first and second language learning, and testing. These contacts led to an increasing awareness of the relevance of statistics for other areas in linguistics and applied linguistics, and of the responsibility of those working in such areas to subject their quantitative data to the same kind of statistical scrutiny as other researchers in the social sciences. In time, students in linguistics made increasing use of the Advisory Service provided by the Department of Applied Statistics at Reading. It soon became clear that the dialogue between statistician and student linguist, if it was to be maximally useful, required an awareness of basic statistical concepts on the student's part. The next step, then, was the setting up of a course in statistics for linguistics students (taught by Woods). This is essentially the book of the course, and reflects our joint views on what linguistics students who want to use statistics with their data need to know.

There are two main differences between this and other introductory textbooks in statistics for linguists. First, the portion of the book devoted to probability and statistical inference is considerable. In order to elucidate the sample–population relationship, we consider in some detail basic notions of probability, of statistical modelling, and (using the normal distribution as an example of a statistical model) of the problem of estimating population values from sample estimates.