Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences E-Book

Paul J. Mitchell

• Herausgeber: John Wiley & Sons
• Kategorie: Wissenschaft und neue Technologien
• Sprache: Englisch

Seitenzahl: 632

Veröffentlichungsjahr: 2022

Beschreibung

Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences

A practical guide to the use of basic principles of experimental design and statistical analysis in pharmacology

Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences provides clear instructions on applying statistical analysis techniques to pharmacological data. Written by an experimental pharmacologist with decades of experience teaching statistics and designing preclinical experiments, this reader-friendly volume explains the variety of statistical tests that researchers require to analyze data and draw correct conclusions.

Detailed, yet accessible, chapters explain how to determine the appropriate statistical tool for a particular type of data, run the statistical test, and analyze and interpret the results. By first introducing basic principles of experimental design and statistical analysis, the author then guides readers through descriptive and inferential statistics, analysis of variance, correlation and regression analysis, general linear modelling, and more. Lastly, throughout the textbook are numerous examples from molecular, cellular, in vitro, and in vivo pharmacology which highlight the importance of rigorous statistical analysis in real-world pharmacological and biomedical research.

This textbook also:

• Describes the rigorous statistical approach needed for publication in scientific journals
• Covers a wide range of statistical concepts and methods, such as standard normal distribution, data confidence intervals, and post hoc and a priori analysis
• Discusses practical aspects of data collection, identification, and presentation
• Features images of the output from common statistical packages, including GraphPad Prism, Invivo Stat, MiniTab and SPSS

Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences is an invaluable reference and guide for undergraduate and graduate students, post-doctoral researchers, and lecturers in pharmacology and allied subjects in the life sciences.

Leseprobe

Experimental Design and Statistical Analysis for Pharmacology and the Biomedical Sciences

Paul J. Mitchell

Department of Pharmacy and Pharmacology University of Bath Bath, UK

Department of Pharmacology and Therapeutics National University of Ireland (NUI) Galway, Ireland

This edition first published 2022© 2022 John Wiley and Sons Ltd

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 law. Advice on how to obtain permission to reuse material from this title is available at http://www.wiley.com/go/permissions.

The right of Paul J. Mitchell to be identified as the author of this work has been asserted in accordance with law.

Registered OfficesJohn Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, USAJohn Wiley & Sons Ltd, The Atrium, Southern Gate, Chichester, West Sussex, PO19 8SQ, UK

Editorial Office9600 Garsington Road, Oxford, OX4 2DQ, UK

For details of our global editorial offices, customer services, and more information about Wiley products visit us at www.wiley.com.

Wiley also publishes its books in a variety of electronic formats and by print‐on‐demand. Some content that appears in standard print versions of this book may not be available in other formats.

Limit of Liability/Disclaimer of Warranty

The contents of this work are intended to further general scientific research, understanding, and discussion only and are not intended and should not be relied upon as recommending or promoting scientific method, diagnosis, or treatment by physicians for any particular patient. In view of ongoing research, equipment modifications, changes in governmental regulations, and the constant flow of information relating to the use of medicines, equipment, and devices, the reader is urged to review and evaluate the information provided in the package insert or instructions for each medicine, equipment, or device for, among other things, any changes in the instructions or indication of usage and for added warnings and precautions. While the publisher and authors have used their best efforts in preparing this work, they make no representations or warranties with respect to the accuracy or completeness of the contents of this work and specifically disclaim all warranties, including without limitation any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives, written sales materials or promotional statements for this work. The fact that an organization, website, or product is referred to in this work as a citation and/or potential source of further information does not mean that the publisher and authors endorse the information or services the organization, website, or product may provide or recommendations it may make. This work is sold with the understanding that the publisher is not engaged in rendering professional services. The advice and strategies contained herein may not be suitable for your situation. You should consult with a specialist where appropriate. Further, readers should be aware that websites listed in this work may have changed or disappeared between when this work was written and when it is read. Neither the publisher nor authors shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages.

Library of Congress Cataloging‐in‐Publication Data

Names: Mitchell, Paul J. (Senior lecturer and Associate Professor in pharmacology) author.Title: Experimental design and statistical analysis for pharmacology and the biomedical sciences / Paul J. Mitchell.Description: Hoboken, NJ : John Wiley & Sons, Inc., 2022. | Includes bibliographical references and index.Identifiers: LCCN 2021034578 (print) | LCCN 2021034579 (ebook) | ISBN 9781119437635 (paperback) | ISBN 9781119437673 (adobe pdf) | ISBN 9781119437666 (epub)Subjects: MESH: Research Design | Pharmacology | Data Interpretation, StatisticalClassification: LCC RM301.12 (print) | LCC RM301.12 (ebook) | NLM QV 20.5 | DDC 615.1--dc23LC record available at https://lccn.loc.gov/2021034578LC ebook record available at https://lccn.loc.gov/2021034579

Cover design: WileyCover image: © RomanOkopny/Getty Images

This book is dedicated to my parents Alec and Jean, long gone but never, ever, forgotten, and to my wife, Angela, and our children Matthew and Samantha of whom I am immensely proud.

Biography

Dr Paul J Mitchell is a Senior Lecturer and an Associate Professor in the Department of Pharmacy and Pharmacology, University of Bath, United Kingdom, and Adjunct Lecturer in the Department of Pharmacology and Therapeutics, National University of Ireland (NUI), Galway, Ireland.

His career in pharmacology started in 1975 when he joined the cardiovascular group led by Dr Bob Poyser at Beecham Pharmaceuticals, Harlow, United Kingdom. After five years, during which he also graduated with an Upper Second BSc in Applied Biology (Pharmacology) from North East London Polytechnic under Prof Geof B West, he transferred to the CNS disorders group led by Dr Mike Clarke. Dr Mitchell left Beecham in 1985 to start post‐graduate studies with Dr (now emeritus Prof) Peter Redfern in the School of Pharmacy and Pharmacology, University of Bath, United Kingdom. After successfully defending his PhD thesis on the Effect of Antidepressant Treatment on Social Behaviour and Circadian Rhythms of Locomotor Activity in the Rat in 1989, he joined the research laboratories of Wyeth‐Ayerst, Taplow, principally to examine the potential antidepressant activity of novel psychotropic compounds in the rodent models of social behaviour that were developed during his postgraduate studies. The results of these behavioural studies were pivotal in the company's decision to fully develop venlafaxine (known in‐house as Wy 45030) to the clinic. Subsequently, Dr Mitchell became heavily involved in the further pre‐clinical development of venlafaxine (Effexor®, Efexor®), the world's first SNRI antidepressant drug, which was approved by the U.S.FDA in 1993 and by the MHRA in the United Kingdom in 1994.

In 1995, Dr Mitchell returned to the University of Bath to set up his own lab to continue examining the effect of antidepressant drugs on rodent social behaviour (Resident‐Intruder test and Social Hierarchy model of social behaviour), while working very closely with the pharmaceutical industry (principally colleagues at Wyeth‐Ayerst in the USA, Lundbeck in Denmark, and Organon in the United Kingdom).

Over the last 25 years, Dr Mitchell has collaborated with colleagues at the University of Bath and NUIGalway to develop a coherent strategy to teach experimental design and statistical analysis to undergraduate and postgraduate students across subject areas in the Life and Biomedical sciences.

Dr Mitchell has been a member of the British Pharmacological Society since 1985. He is currently working closely with the society on a residential training workshop on the topic of this book covering the principles of robust, rigorous, experimental design, and statistical analysis. This course is ideal for early career researchers working in drug discovery or academia.

Acknowledgements

Homo Sapiens – Part 1

Looking back over copious notes and draft versions of this book, I've come to realise that this project has taken me far longer than my initial 12‐month plan envisaged. But let me take you back to the beginning of the 2016–2017 academic year when I confronted a group of pharmacology undergraduate students who had just returned from a year's placement and were about to embark on the final year of their degree programme. My question ‘So, how're your stats skills coming along now you've been in the big wide world?’ was met with a mixture of dismay and disdain, not to say contempt for mentioning the S word! The ensuing conversation quickly dispelled any thoughts on my part regarding their progress and so I quickly put together an intense four‐week package, which covered all my statistics lectures and hands‐on data analysis workshops to get them up to speed. The feedback I received suggested that delivering such material in such a short period was successful, but one important component was missing; they had no access to a suitable concise textbook covering descriptive and inferential statistics suitable for undergraduate pharmacology students. Indeed their major complaint was that currently available textbooks were either focused on statistical theory (suitable only for highly competent students of mathematics) or were simply user guides for statistical software packages; both of which were totally inappropriate for students who simply wanted to know which statistical tests they had to use to analyse different types of experimental data. In short, they needed a concise ‘textbook’ book, which showed them how to use statistics as a tool with which they could analyse their experimental data and arrive at appropriate conclusions, thereby revealing the relationship of their data to the real world. So, it seems obvious to me that the first group of individuals that I should thank, and you, dear reader, should blame for kick‐starting this project which has resulted in the contents of this book, are the infamous seven final year pharmacology students from 2016: Charlotte Bell, Charlotte Day, Sam Groom, James Miles (yes, that cocky bugger), Katy Murrell, Gemma Wilkinson, and Alex Williams – according to my latest information most if not all have now completed (or nearly so) their subsequent PhD studies and are now forging research careers on their own (and most notably without my help!). If this book is in any way successful, then clearly you seven should be held totally accountable!

Of course, when you embark on a project that is clearly not your day job, then you need a lot of support to help you to find the time in the working day that enables you to turn that germ of an idea into fruition. My sincere thanks to all my colleagues in the Department of Pharmacy and Pharmacology at the University of Bath that in many, diverse, ways have enabled me to focus on putting this manuscript together, that have humoured me while I've ranted on regarding the statistical inadequacies in the scientific press or listened quietly while I've bombarded them with different ideas on how to describe quite complex statistical issues that an inexperienced undergraduate student may (hopefully) understand. Most importantly, my thanks to Profs Steve Ward and Roland Jones who agreed for me to move my teaching duties around that created gaps in my teaching load, which allowed me to concentrate on writing. This book is full of data examples, which, I hope, will enable the reader to understand more‐fully descriptive and inferential statistics and to envisage statistics in action. Most of the data examples are my own, and for all other examples I am very grateful to Dr Malcolm Watson. I must also thank Prof Steve Husbands and Dr Christine Edmead for their helpful comments, encouragement, and suggestions after reading the first completed draft version of the manuscript; of course, considering Steve is a medicinal/organic chemist, who (by his own admission) failed to understand anything described in the book, his comments were totally ignored!

For the last 15–20 years or so, I have worked very closely with Prof John Kelly, Department of Pharmacology and Therapeutics, NUIGalway, during which we have tried, successfully it must be said, to develop a fully integrated series of lectures and workshops to teach undergraduate and postgraduate students in pharmacology, neuropharmacology, toxicology, and drug discovery the vagaries of robust experimental design and statistical analysis. I still travel to Galway every year to expose John's postgraduate students to an English sense of humour in my attempt to run hands‐on statistical workshops – I must be doing OK as John keeps inviting me back! My sincere thanks to John for all the support and encouragement he has given me during that time and to other colleagues in Galway, notably Ambrose O’Halloran and Sandra O'Brien, who were instrumental in preparing the initial versions of the lectures, which I now subject my own students to back in Bath and which are closely aligned to the contents and flavour of this book.

During the life of this project, I have worked closely at various times with the management team for the British Pharmacological Society and I would like to convey my thanks to David James (Executive Director, Business Development) for his initial help and advice to get this project off the ground, and latterly to Katherine Wilson (Director, Research Dissemination) and Lee Page (Head of Education and Engagement) for their help and advice on ‘what to do next’!

I am also very grateful to everybody at Wiley from Alison Oliver, who as Publications Manager and Commissioning Editor back in 2016 took a risk and encouraged me to put my ideas into a proposal, which subsequently became a formal agreement between myself and Wiley, to James Watson (Publications Manager), Kimberly Monroe‐Hill (Managing Editor), and Tom Marriott (Assistant Editor, Health and Life Sciences) who have guided me through all the steps following formal submission of the final manuscript through to publication. My thanks also to the reviewers of my initial proposal to Wiley who thought this project was a good idea, who have encouraged me to complete the project ever since (hi John and Steve, you know who you are) and who also opened my eyes that this work may be not only useful within the realm of pharmacology but also throughout the biomedical and life sciences!

My career in pharmacology has taken me from the pharmaceutical industry with Beecham Pharmaceuticals in the 1970s and 1980s, through my PhD studies at the University of Bath in the latter half of the 1980s (under the invaluable supervision of Prof Peter Redfern), then back to the industry with Wyeth‐Ayerst in 1989 before returning to the University of Bath in 1995 where I have remained ever since. Throughout that time in industry and academia, I have worked with a wide range of wonderful, highly skilled, individuals and made life‐long friends too numerous to name individually here (but you should all know who you are on both sides of the Atlantic Ocean). I shall remain eternally grateful for all your help, guidance, encouragement, patience, comments, critique (usually constructive), and tutelage throughout my career in pharmacology.

Statistical Packages

You will note that at the end of most of the chapters in this book, I have been able to provide screenshots from the software packages that I used to analyse the examples used in the book. This was for two purposes. First, this allowed me to check my own calculations for every single example and statistics test described herein (yes; every example has been analysed in the good old‐fashioned way by hand and a good calculator – good God, what a geek I hear you cry – and you'd be right as my academic colleagues keep telling me!), and second, I hope that when you run your own data analysis you will now be forewarned about what to expect (and not be surprised) by the output from the software you have used. To that end, I am most grateful to the software companies concerned for permission to reproduce screenshots from their software. Consequently, screenshots from GraphPad – Prism®Statistics software version 8.2 and above are printed with permission of GraphPad Software, San Diego, California, USA; screenshots from MiniTab software version 18 and above are printed with permission of MiniTab, LLC; screenshots from InVivoStat are reprinted with permission of the InVivoStat team (specifically Simon Bate); and finally screenshots from IBM® SPSS® Statistics software (SPSS) version 26 and above are printed with permission from International Business Machines Corporation (IBM).

Homo Sapiens – Part 2

I've been very lucky to have a number of very close friends who have remained loyal regardless of where my career has taken me, so a special mention to Dave Bragg (‘Braggy’), John Clapham (JC), and Alan Rainbird for your unwavering friendship (which I value more than words can ever express) since we first met over 45 years ago, John Kelly (see above), and more recently to Kevin McDermott for dragging me out most Saturday mornings to the golf course to clear our heads for a few therapeutic hours away from the stress of our professional lives, see you on the first tee mate!

Finally (!), all my love to my wife Angela and my children Matthew and Samantha – how you all ever put up with such a cantankerous old git as myself (especially during the last four years or so while I worked on this manuscript) I shall never know. You are all my rock, and I will always be forever grateful.

1Introduction

Experimental design: the important decision about statistical analysis

Whenever you make plans for your annual holiday, you do not just pack your suitcase willy‐nilly without first making plans about what you want to do, where you want to go, how you are going to get there, etc. For example, if your idea is to go trekking around the coast of Iceland, then you would look really stupid if, on arrival in Reykjavik, you opened your suitcase only to find beachwear and towels! Indeed, identifying what you want to do on holiday and where you intend to go determines what you need to take with you and what travel arrangements you need to make. In fact, what you do on holiday can be viewed as the final output of your holiday arrangements. The same can be said for the design of any well‐planned, robust, scientific experiment. The final output of your experiment, i.e. the communication of your results, whether it be a figure (scatter graph, bar chart, etc) or table, largely determines every single step in the preceding experimental design, including the strategy of your statistical analysis.

Figure 1.1