Statistics Toolkit E-Book

Rafael Perera

• Herausgeber: John Wiley & Sons
• Kategorie: Fachliteratur
• Serie: EBMT-EBM Toolkit Series
• Sprache: Englisch

Seitenzahl: 121

Veröffentlichungsjahr: 2011

Beschreibung

This concise book will help you to interpret the statistical evidence provided by quantitative studies and to plan how to work with data in your own clinical research. Following the successful format of the Toolkit series, Statistics Toolkit guides the reader through statistical concepts using flowcharts, diagrams and real life examples to reflect concepts in a simple and practical manner. Sections include: * * Clear explanation of basic concepts in the context of clinical research * Demonstration of how data are described, displayed and interpreted in different formats * Practical glossary and key to the symbols used in statistics and a discussion of the software tools The book offers a handy, quick reference that has an easy-to-follow structure throughout, making it ideal for health care professionals and students.

Leseprobe

Contents

Introduction

Data: describing and displaying

Summarizing your data

The arithmetic mean (numeric data)

The geometric mean

The weighted mean

The median and mode

Measures of dispersion: the range

The variance

The standard deviation

Percentiles

Standard error of the mean

Displaying data

Probability and confidence intervals

Probability distributions

Confidence intervals

Hypothesis testing

The steps in testing a hypothesis

Errors in hypothesis testing

Statistical power

Randomised controlled trials: mean, median, mode, RD, NNT, Mann-Whitney and log rank tests

The mean and the median

The risk difference (RD) and the number needed to treat (NNT)

The Mann–Whitney test

Log-rank test and Kaplan–Meier survival statistics

Systematic reviews 1: odds, odds ratio, heterogeneity, funnel plots

Odds ratios

Combining odds ratios

Heterogeneity tests

Visualising statistical significance

Subgroup analysis

Testing for publication bias

Other terminology used

Case-control studies: odds ratios, logistic regression

Odds ratios in case-control studies

Logistic regression and adjustment for crude odds ratios

Questionnaire studies I: weighted mean frequency, nonparametric tests

Weighted mean frequency

Nonparametric bootstrapping was done to calculate 95% CIs

The median time of onset and interquartile range

Questionnaire studies 2: inter-rater agreement

Likert scales

Assessing inter-rater agreement using kappa

Cohort studies: prevalence, incidence, risk, survival, mortality rate, association, prognostic factor

Using mortality or survival rates

Regression: are the outcomes associated with prognostic factors?

Interpreting multivariable analysis

Systematic reviews 2: Cohort study odds ratios and relative risk

Hazard ratio

Cox proportional hazards model

Diagnostic tests: sensitivity, specificity, likelihood ratios, ROC curves, pre- and post-test odds, pre- and post-test probability

How well does the test detect the disease?

How likely is a given test result to be true?

Likelihood ratios (LR)

What about tests that have a range of possible results?

Measuring test performance at different cut-off points

Statistical significance

Pre-test odds and post-test odds: particularising the evidence

Using Bayes theorem and the Fagan nomogram

Scale validation: correlation

Visual analogue scale and CARIFS

Correlation coefficient (Pearson and Spearman)

Factor analysis and Cronbach’s alpha

Statistical toolkit: Glossary

Software for data management and statistical analysis

References

Index

© 2008 Rafael Perera, Carl Heneghan and Douglas Badenoch

Published by Blackwell Publishing

BMJ Books is an imprint of the BMJ Publishing Group Limited, used under licence

Blackwell Publishing, Inc., 350 Main Street, Malden, Massachusetts 02148-5020, USA

Blackwell Publishing Ltd, 9600 Garsington Road, Oxford OX4 2DQ, UK

Blackwell Publishing Asia Pty Ltd, 550 Swanston Street, Carlton, Victoria 3053, Australia

The right of the Author to be identified as the Author of this Work has been asserted in accordance with the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, except as permitted by the UK Copyright, Designs and Patents Act 1988, without the prior permission of the publisher.

First published 2008

1 2008

ISBN: 978-1-4051-6142-8

A catalogue record for this title is available from the British Library and the Library of Congress.

Set in Helvetica Medium 7.75/9.75 by Sparks, Oxford - www.sparks.co.uk Printed and bound in Singapore by Markono Print Media Pte Ltd

Commissioning Editor: Mary Banks

Development Editors: Lauren Brindley and Victoria Pittman

Production Controller: Rachel Edwards

For further information on Blackwell Publishing, visit our website: http://www.blackwellpublishing.com

The publisher’s policy is to use permanent paper from mills that operate a sustainable forestry policy, and which has been manufactured from pulp processed using acid-free and elementary chlorine-free practices. Furthermore, the publisher ensures that the text paper and cover board used have met acceptable environmental accreditation standards.

Designations used by companies to distinguish their products are often claimed as trademarks. All brand names and product names used in this book are trade names, service marks, trademarks or registered trademarks of their respective owners. The Publisher is not associated with any product or vendor mentioned in this book.

This handbook was compiled by Rafael Perera, Carl Heneghan and Douglas Badenoch. We would like to thank all those people who have had input to our work over the years, particularly Paul Glasziou and Olive Goddard from the Centre of Evidence-Based Medicine. In addition, we thank the people we work with from the Department of Primary Health Care, University of Oxford, whose work we have used to illustrate the statistical principles in this book. We would also like to thank Lara and Katie for their drawings.

Introduction

This ‘toolkit’ is the second in our series and is aimed as a summary of the key concepts needed to get started with statistics in healthcare.

Often, people find statistical concepts hard to understand and apply. If this rings true with you, this book should allow you to start using such concepts with confidence for the first time. Once you have understood the principles in this book you should be at the point where you can understand and interpret statistics, and start to deploy them effectively in your own research projects.

The book is laid out in three main sections: the first deals with the basic nuts and bolts of describing, displaying and handling your data, considering which test to use and testing for statistical significance. The second section shows how statistics is used in a range of scientific papers. The final section contains the glossary, a key to the symbols used in statistics and a discussion of the software tools that can make your life using statistics easier.

Occasionally you will see the GO icon on the right. This means the difficult concept being discussed is beyond the scope of this textbook. If you need more information on this point you can either refer to the text cited or discuss the problem with a statistician.