Quantitative research in linguistics: An introduction (review)

Rasinger's new book provides a comprehensive and accessible introduction to quantitative research methods in linguistics. Aimed at an advanced undergraduate and early postgraduate audience, the book addresses a longstanding need in the field for a textbook specifically geared toward students who may have already developed a nuanced understanding of language and linguistics but who have had little to no exposure to statistical methods. As such, the book stands to make an important contribution to the teaching of linguistics, whether as a classroom resource for teachers or as a guide to individual students as they undertake their own research projects.

The book is divided into eight chapters (plus an introduction and an appendix) grouped into two parts. Part 1 (Chs. 2–4) is devoted to a discussion of various aspects of data collection. It begins in Ch. 2 (9–34) with a somewhat long, yet in many ways unavoidable, inventory of core concepts in quantitative research, including (in order) deductive versus inductive reasoning, dependent and independent variables and variable operationalization, causality and latency, levels of measurement (nominal, ordinal, interval, and ratio data are all described), continuous versus discrete data, reliability and validity, and hypothesis testing. The definitions are clearly and concisely presented, and, thanks to the extensive use of examples, R succeeds in making concrete what are at times rather abstract concepts. I wonder, though, whether the order of presentation of these concepts is as intuitive as it could be. Reliability and validity, for example, are normally discussed in relation to deductive and inductive reasoning, while here they are tacked onto the end of a discussion of different data types. Similarly, an explanation of the difference between continuous and discrete data would logically seem to precede a discussion of interval versus nominal measurements, while here it appears afterwards. Finally, and on a slightly different note, I was surprised to see no mention of null hypotheses in the section on hypothesis testing (though a discussion of null hypotheses does appear later in Part 2). Yet, these minor concerns notwithstanding, I believe that Ch. 2 provides a strong theoretical foundation for the methods taught throughout the rest of the book.

Following on from this more general introduction in Ch. 2, Ch. 3 (35–55) turns to the issue of research design and sampling. Here, R does an excellent job of laying out the different design types, including cross-sectional designs, longitudinal designs (both panels and cohorts), and experimental designs. There is also a brief yet incredibly clear description of apparent versus real time and of the benefits and drawbacks of both synchronic and diachronic approaches. The chapter then closes with a discussion of sampling and of the crucial yet unfortunately often overlooked distinction between samples and populations.

Ch. 4 (56–83) is devoted entirely to a discussion of questionnaires. Under the heading 'know what you want and ask what you need', R describes the various issues involved in questionnaire design, moving from a discussion of the relative merits of open versus closed questions to a discussion of different response set types. This is followed by a detailed exposition of the various layout options that exist, with the so-called 'best' layout illustrated in a sample questionnaire at the end of the chapter. Finally, R moves on to a discussion of coding and the different ways in which questionnaire data can be organized in preparation for quantitative analysis. And while I certainly see the logic in placing this discussion of coding in the chapter on questionnaires, it is nevertheless unfortunate that its inclusion here (and nowhere else in Part 1) makes it appear as if coding is something that you do only when your data are derived from questionnaires. A short mention of the necessity for coding no matter what the data-collection method employed (either in Ch. 4 or in Ch. 3) would have done a great deal to alleviate this potential source of confusion.

Part 2 of the book (Chs. 5–9) shifts focus from data collection to quantitative...