This article is only available in the PDF format. Download the PDF to view the article, as well as its associated figures and tables.
Surveys by several authors suggest that the number of poorly designed and poorly analyzed studies that are published in reputable medical journals is distressingly high. As a group, researchers must thus be more fully aware of the importance of the statistical aspects of a study (which may include consultation with statisticians) in order that conclusions are justifiably drawn. Readers also must possess the tools to assess critically whether a study is based on correct and relevant statistical principles, since the appearance of a study in a journal does not necessarily imply statistical accuracy. Statistics in Practice seeks to provide a framework for nonstatisticians to upgrade their statistical understanding. The book succeeds admirably.
The first section of the book, by Mr Altman, is subtitled "Statistics and Ethics in Medical Research" and is organized into eight chapters. He begins with the well-argued proposition that bad statistics is unethical. The more usual concern
Lubin JH. Statistics in Practice. JAMA. 1982;248(13):1647. doi:10.1001/jama.1982.03330130095047