**PREFACE**

**Concise Encylopedia of Biostatistics for Medical Professionals**

**By**

**Abhaya Indrayan and Martin Holt**

The science of biostatistics has enormously proliferated during the past few decades and its medical applications seem to have exponentially increased. Our interaction with a large number of medical practitioners and researchers reveal that they increasingly get perplexed over biostatistical methodology and terminology used in medical literature. Understanding biostatistical concepts is sometimes a challenge for medical professionals because they often lack formal training in this area. Encyclopedias are commonly used as reference to quickly learn the essentials of a specific topic. We believe that our mini-encyclopedia (

*ency*) will be a handy and often-used resource for anyone needing to quickly understand the basics of a particular biostatistical concept with a minimal level of mathematical knowledge. The topics, including alternative phrases, are arranged alphabetically, and the reader can easily locate the topic of his or her interest and understand the meaning of the terms involved, their applicability to health sciences, and their medical interpretation. We have tried to orient the discussion of the topics firmly to the needs and backgrounds of medical and health professionals so that it resonates with them. Like any encyclopedia, this one too is not a replacement for a textbook as the matter is not arranged in a logical sequence. For a regular book in similar mode, see the latest edition of Medical Biostatistics by Abhaya Indrayan (CRC Press). This book takes you through most of these topics in a logical flow beginning to end using medical uncertainties as the common theme for navigation.

Our target audience is those medical professionals who want to adhere to an evidence-based approach to medical practice. This requires good understanding of biostatistical methods for proper interpretation of results reported in scientific articles. These professionals include physicians, surgeons, dentists, pharmacists, nursing professionals, nutritionists, and epidemiologists. Academia, especially graduate students in medical and health sciences, may also find this

*ency*useful as a reference.

We have conceived this encyclopedia as relatively short, hence the prefix “Concise”. It focuses on conceptual knowledge and practical advice in easy-to-read format rather than mathematical details. This approach hopefully enhances the book’s usefulness as a reference for medical and health professionals. Limiting the book to a single volume will also make it more affordable and attractive to individual professionals and not just libraries. Because of the limited size, the explanation provided for various terms in this volume is and cannot be comprehensive, and some readers may feel that the kinds of details they need are missing. For some terms such as regression and analysis of variance, full chapter in many books and in fact full books are available. Our explanation is limited to the essential features, applications and interpretations we perceive as relevant to medical sciences. However, in this process, we may not have been able to meet the exacting standards of some statisticians.

It is not possible to cover everything in one volume yet this

*ency*may be quite comprehensive for the target audience. It defines and explains more than 1000 commonly and not so commonly used key biostatistical terms and methods ranging from simple terms such as mean and median to advanced terms such as multilevel models and generalized estimating equations. Those health topics that have significant biostatical component are also presented. These include terms relating to community health, and social and mental health also.

We have listed alternative terms to make search easier and also to help our medical colleagues in removing some of the confusion about multiple terms for the same method or topic. We have also tried to explain how they are related, if at all. For example, an independent variable in a regression setup is also called regressor, an explanatory variable, a predictor, a covariate, a confounder, a concomitant variable, etc., depending upon the context. Similarly a dependent variable is also called a response, a target or an outcome variable. We discuss all of them together under one unifying topic ‘dependent and independent variables’ so that their subtle distinction also can be explained. All the related terms for each topic are also listed for completeness but refer the reader to the unifying topic. For these few terms, the length of the section has considerably increased. On the other hand, a topic that needs to be extensively described with varied applications, such as chi-square, is split into several relatively smaller topics to retain the focus and to keep each topic within a manageable limit. The purpose is to remain short and crisp for the convenience of our medical readers. For some topics, such a split may have hindered in providing a correct perspective. Also, we have tried to describe each topic in a self-contained manner for independent reading so that the reader does not have to flip through back and forth although some cross-referencing for the involved terms looked unavoidable. Thus there is some duplication that seems inevitable in this kind of work although we have tried to minimize this, sometimes by combining two related topics.

Given that the target audience is medical professionals, formulas and mathematical details are limited to high school algebra. Thus formulas for only those terms that do not require complicated mathematics have been provided. For others, a heuristic approach is adopted to explain and describe the rationale and applications. We are aware that heuristics can generate imprecision and sometimes even errors – thus imperative care has been taken. Many topics are illustrated with the help of medical examples and figures, including references to contemporary medical literature. The list of references for topics is not comprehensive though, and is restricted to the most relevant ones, so that the reader does not feel burdened. Weblinks have been provided for references wherever we could locate so that the reader can directly access the referenced material. This could not be done for the referenced books. The date of last access by us is written only for those websites that are likely to change. For ostensibly permanent websites such as of journals, the date of last access is not provided. These links provide an indispensable platform for further study and research.

We realized the difficulty in communicating complex mathematical issues in a simple nonmathematical language and accepted this challenge. Efforts all through this

*ency*are on simple language that can be appreciated by the target audience who have much less grinding in mathematics. We have not used formal language, instead the format is conversational. In the process though, we may not have been very accurate in explaining a method or a concept. This admittedly is a limitation of this book. But our focus is firmly on medical applications and concerns. For intricate methods, where we have explained only the concept and applications but not the analysis, a reference is provided wherein one can find the required details. Historical briefs, biographical snippets, and photographs of prominent statisticians are added to provide a broader context for the topics covered.

Within each topic, those terms that are explained elsewhere in this volume are in bold when they appear first, particularly when details of that term is likely to help the reader to understand the current topic. There might be some minor variation in the topic title as required for the current text. Other statistical terms appear in italics and separately listed so that these can also be located. Phrases that we want to emphasize also are in italics.

**To those of our medical colleagues who are not familiar with statistical notations, we advise they see the topic ‘notations’ so that they feel comfortable in reading and understanding this**We have tried to follow a uniform system of notations across the topics.

*ency*.Almost all other encyclopedias are contributed by several authors and edited by experts. Perhaps ours is the first attempt to prepare an encyclopedia by two authors. This has advantages and disadvantages. Advantage is uniformity of level of discussion—when based on contributions by different authors, some topics attain too high a level and others a too low level depending on the authors’ understanding of the topic. The second advantage is our consistent style. A disadvantage is that this may not go in depth of some topics that topic experts would prefer.

Despite best care, errors may have occurred. We are sure that the book will be critically examined by the reviewers. Their feedback will help in improving the next edition.

**Abhaya Indrayan**

**Martin P. Holt**

First author's website: http://indrayan.weebly.com