Numbers can justify library activities: a book review

Leo Egghe, Ronald Rousseau   Elementary Statistics for Library and Information Service Management. London: Aslib-IMI, 2001.  120 p. ISBN 0-85142-451  £24.99

"The story of millions" probably is the most successful phrase representing library circulation that I have ever heard in interpreting library statistics. It was a comment on the annual report of Helsinki City Library in early 1990s. Then the annual report of a library was a rare event, if we do not count reports that were only for internal official use. Now, in the first decade of the new millennium, annual reports are an obligatory activity for library managers as, more often than ever before, the funding authorities are requesting rationale for money. Transparency is another reason for constant dissemination of information on library activities to the community. Correct statistical data are the basis for an annual or any other report if we are concerned about the legitimacy of our operations.

Among other skills and knowledge, an effective manager should master statistical procedures for data collection and interpretation. Statistics is defined as information based on a study of the number of times something happens or is present, or other numerical facts, it also is the science of using information discovered from studying numbers (The Cambridge International Dictionary of English.) . In management, statistics is used for reporting, arguing for the decisions, and forecasting the future changes. So, the book "Elementary Statistics for Effective Library and Information Management" is exactly what those who are looking for a short and simple enough text on library statistics for managers are looking for. However, the reader of this book should have some level of awareness in statistics. If you are not one of them, before starting to read, equip yourself with a dictionary of statistical definitions or, as the authors recommend, "have access to a statistical software package or at least to a pocket calculator with some statistical functions" Having the necessary tools be prepared to allocate time for reasonable study. The book is structured into three major parts – Information about the library, Descriptive statistics and Inferential statistics. The appendices supplement the text and are of great use to the reader.

The first part of the book covers managerial aspects, gathering of data, sampling methods. First of all, the authors describe different audiences of reports. The authors are to be complimented on shattering the myth that computers make the collection of data easier. Accuracy and exact definitions of counted attributes are what one should have in mind. The simplicity of the description of sampling methods is another good aspect of this first section of the book.

The second part deals with the presentation of raw data in a graphical way. Bar diagrams, histograms, problems of graphical representations, scatter plots and regression lines are just a few topics discussed here. Managers shall be interested in Lorenz curve, a graphical tool to find cause and effect relationships. It is explained using the popular 80/20 rule, stating that a relatively small group determines a much larger part. Under the sections you will find exercises, so that the acquired information can be applied immediately. Unless you finished high school, or its equivalent, be prepared to recover mathematical knowledge with the professional guidance of the authors.

Anyone who has been involved in quantitive research, at some moment raises the question of representability. How to make conclusions on totality with partial data? How many samples and sets of measurements one needs for reliable conclusions? To answer - open the third part Inferential statistics which discusses precisely what you need - the conclusions on samples. Among the explanations by means of examples the reader will find arguments of the possibility of one-sided tests, discussion on binominal fractions and sample sizes for multinomial fractions. Use of simple or more spectacular methods of inferential statistics, described in the book, enable us to make correct or calculated decision.

If you reach the end of the book it will ensure that you have acquired confidence and can report or make decisions based on the manipulation of statistical data. It definitely will make your report more professional and your proposals difficult to criticize. However, do not forget that sometimes in real life leaders should be prepared to manage by guessing because yesterday's facts, might not be true tomorrow.

Ramune Petuchovaite
Vilnius University
August 2001