Title Reviewed:
An Introduction to Quantitative Methods for Historians

Author Reviewed:
Roderick Floud

Author:
Edwin R. Coover

Date:
1974

Source:
Indiana Magazine of History, Volume 70, Issue 1, pp 88-89

Article Type:
Book Review

Download Source:
xml

An Introduction to Quantitative Methods for Historians. By Roderick Floud. (Princeton, N. J.: Princeton University Press, 1973. Pp. ix, 220. Notes, tables, figures, bibliography, index. $7.50.)

Roderick Floud, an assistant lecturer in economic history at Cambridge University, has added another volume to the fast growing group of quantitative methods "cookbooks" for historians. His volume presents some attractive features. Comparatively, it is inexpensive and brief. It is also relatively free of errors. Professor Floud writes well and is a good explainer. He employs historical examples throughout, and his chapter six, "The Analysis of Time Series," is an unusually strong treatment.

Nevertheless, the volume contains serious drawbacks for historians seeking an introduction to quantitative methods. The principal one surrounds Floud's pronouncement on ordinal data (p. 43): "ordinal data are rarely encountered in historical work, and it is therefore unnecessary to consider examples of them at length." Later (p. 128), he retreats from this stand somewhat: "It should be noted, however, that it is sometimes sensible to make use of methods suitable for ordinal data, when the data are apparently of interval type, but when one has some doubts about them." He ignores the frequent occurrence where the few variables that are of interval level are grouped to ordinal to enable a cross classification analysis with the remainder that are nominal or ordinal. His view on the infrequency of ordinal data accounts for the omission of any general discussion of non-parametric statistics and, in this reviewer's opinion, seriously compromises the volume's usefulness for the beginner in the field.

The other shortcomings of the volume are less fundamental but also detract from its usefulness. One suspects that his last chapter, "Computers and Data Processing Equipment," reflects less on Floud's knowledge of the subject than it does on the contrasts between data processing environments here and abroad. Simply, by the standards of most state universities in the United States, his description of data processing is ten years behind the times. There is no mention, for instance, of the modern social data analysis systems such as SPSS, DATA-TEXT, OSIRIS II, P-STAT, etc., which have so facilitated data handling and analysis in recent years. Finally, some may find several aspects of the volume's organization somewhat peculiar. A discussion of samples and populations is delayed to the next to last chapter. Also saved for this late chapter, titled "The Problem of Imperfect Data," is a brief treatment of theory and models. Organization is a matter of individual preferment, but rarely will the historian encounter problems of theory, sampling, and imperfect data late in his analysis. In sum, there are better books for the beginner in this area.

Indiana University, Bloomington

Edwin R. Coover



Published by theĀ Indiana University Department of History.