Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Statistics with Confidence: Confidence Intervals and Statistical Guidelines

Statistics with Confidence: Confidence Intervals and Statistical Guidelines REVIEWS (Part 2, multiple-factor logistic ~od~ls. and the tools p~oper of mUltivari~te analysis are not menti?ned. When discussing causality It IS necessary to mtroduce the notion that a group of vanables may be influenced by a factor while the individual variables within that group may not appear to show any meaningful change. Thus the concept as defined by the set is influenced by the factors while the individual constituents-the observables-may not seem to be. In the analysis of causality as with the presentation of any kind of data, one seeks for simplicity of explanation without oversimplification-a delicate balance. The greater the emphasis of fundamentals at this stage the less likely there are to be misunderstandings when the methods used become more complex. I commend this book to the discerning data analyst and to all those involved in the analysis of medical trials. Colin P. Chalmers Birkbeck College London 2. Statistics with Confidence: Confidence Intervals and Statistical Guidelines. Edited by M. J. Gardner and D. G. Altman. ISBN 0 7279 0222 9. British Medical Journal, London, 1989. 140 pp. £7 .95. This book is an edited and revised collection of papers originally published in the British Medical Journal. The editors jointly wrote five of the 10 chapters and Martin Gardner has contributed to all the other chapters and is responsible for the tables. The main part of the book is on estimation and confidence intervals. The last two chapters, written with medical studies and journals in mind, are on statistical guidelines and checklists. Each chapter is relatively self-contained, but there is an underlying thread of continuity running through the book. The book starts by discussing the change in emphasis from hypothesis testing to confidence intervals in presenting results. It continues with clearly written chapters on calculating confidence intervals. An expression for an interval is followed immediately by a worked example using real data from medical studies. In addition to well-known confidence intervals for means, proportions, regression and correlation, confidence intervals for relative risks, odds ratios, standardized ratios and rates, survival time analyses, and quantiles are covered. The tables are in a useful format for calculation of confidence intervals. There are no theoretical derivations, but references are given for those who want to know more. In all, the book is useful to have on one's shelves. The statistical background expected of the reader is unclear. For instance, there is a fairly detailed account of what confidence intervals are, but familiarity with hypothesis testing seems to be assumed. This is possibly an indication of the state of statistical knowledge in the medical profession, but makes the book less suitable for general recommendation. The term standard deviation refers to samples, but is not defined algebraically, a serious omission as there are many who divide by the sample size when finding the variance. There is no mention of the normal approximation to the t-distribution, perhaps because the t-tables provided give degrees of freedom from I to 400 inclusive. Computer software for IBM or compatible microcomputers has been written as a companion to the book. I was sent a test disc and a draft of the manual. The manual was clear and gave screen output for a large number of worked examples from the book. The software was very easy to use. My main criticism is that there seemed to be no way within the program of changing an option chosen in error, and I found the wait before the welcome message appeared rather long. F. R. Jolliffe BruneI University Uxbridge http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of the Royal Statistical Society Series A (Statistics in Society) Oxford University Press

Statistics with Confidence: Confidence Intervals and Statistical Guidelines

Download PDF
1 page

 
/lp/oxford-university-press/statistics-with-confidence-confidence-intervals-and-statistical-toOgx09NUO

References (0)

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Copyright
© 1990 Royal Statistical Society
ISSN
0964-1998
eISSN
1467-985X
DOI
10.2307/2982808
Publisher site
See Article on Publisher Site

Abstract

REVIEWS (Part 2, multiple-factor logistic ~od~ls. and the tools p~oper of mUltivari~te analysis are not menti?ned. When discussing causality It IS necessary to mtroduce the notion that a group of vanables may be influenced by a factor while the individual variables within that group may not appear to show any meaningful change. Thus the concept as defined by the set is influenced by the factors while the individual constituents-the observables-may not seem to be. In the analysis of causality as with the presentation of any kind of data, one seeks for simplicity of explanation without oversimplification-a delicate balance. The greater the emphasis of fundamentals at this stage the less likely there are to be misunderstandings when the methods used become more complex. I commend this book to the discerning data analyst and to all those involved in the analysis of medical trials. Colin P. Chalmers Birkbeck College London 2. Statistics with Confidence: Confidence Intervals and Statistical Guidelines. Edited by M. J. Gardner and D. G. Altman. ISBN 0 7279 0222 9. British Medical Journal, London, 1989. 140 pp. £7 .95. This book is an edited and revised collection of papers originally published in the British Medical Journal. The editors jointly wrote five of the 10 chapters and Martin Gardner has contributed to all the other chapters and is responsible for the tables. The main part of the book is on estimation and confidence intervals. The last two chapters, written with medical studies and journals in mind, are on statistical guidelines and checklists. Each chapter is relatively self-contained, but there is an underlying thread of continuity running through the book. The book starts by discussing the change in emphasis from hypothesis testing to confidence intervals in presenting results. It continues with clearly written chapters on calculating confidence intervals. An expression for an interval is followed immediately by a worked example using real data from medical studies. In addition to well-known confidence intervals for means, proportions, regression and correlation, confidence intervals for relative risks, odds ratios, standardized ratios and rates, survival time analyses, and quantiles are covered. The tables are in a useful format for calculation of confidence intervals. There are no theoretical derivations, but references are given for those who want to know more. In all, the book is useful to have on one's shelves. The statistical background expected of the reader is unclear. For instance, there is a fairly detailed account of what confidence intervals are, but familiarity with hypothesis testing seems to be assumed. This is possibly an indication of the state of statistical knowledge in the medical profession, but makes the book less suitable for general recommendation. The term standard deviation refers to samples, but is not defined algebraically, a serious omission as there are many who divide by the sample size when finding the variance. There is no mention of the normal approximation to the t-distribution, perhaps because the t-tables provided give degrees of freedom from I to 400 inclusive. Computer software for IBM or compatible microcomputers has been written as a companion to the book. I was sent a test disc and a draft of the manual. The manual was clear and gave screen output for a large number of worked examples from the book. The software was very easy to use. My main criticism is that there seemed to be no way within the program of changing an option chosen in error, and I found the wait before the welcome message appeared rather long. F. R. Jolliffe BruneI University Uxbridge

Journal

Journal of the Royal Statistical Society Series A (Statistics in Society) – Oxford University Press

Published: Dec 5, 2018

There are no references for this article.