spacer
home > epc > spring 2013 > a methodological approach
PUBLICATIONS
European Pharmaceutical Contractor

A Methodological Approach

In his recent book, Thinking, Fast and Slow, Daniel Kahneman discusses how statisticians make intuitive judgements about experiments they are analysing. He and his late colleague Amos Tversky concluded that statisticians are not good with intuition. They found that they greatly exaggerate the likelihood that the results of an experiment will be successfully replicated, even with a small sample, and that they even gave very poor advice to a graduate student in their experiment, about the numbers of observations that she would have to collect to conduct her (fictitious) study. This work, which is reported in full in the book, was the starting point for Kahneman to begin the work with Tversky which eventually won him the Nobel Prize.

Statistics in Medicine is a weighty tome. Now in its third edition, it has been fully revised and would seem to provide an excellent pathway for people wishing to learn about statistical methods in medicine. The book covers many basic statistical techniques and provides a good starting point for further enquiry. It also gives an excellent piece of advice: budding statisticians should meet qualified statisticians and ask questions about their methodology and their experience. The book is careful to distinguish between observational studies and clinical trials (experimental studies) and offers wise advice that observational studies should make modest claims; advice which should be heeded by the modern media who trumpet the results of such studies on the front pages of our newspapers. This, of course, links back to the observations of Kahneman and Tversky regarding intuitive statisticians.

This work provides a thorough grounding in the statistical methodology that a budding statistician would wish to employ. It claims to cover 90 per cent of all statistical analyses that are likely to be encountered. Where a topic is not treated in great depth, for example Bayesian analysis, useful references are sited for further enquiry.

Many examples of statistical analyses are given in the book and these are taken from real medical studies, although some are edited better to illustrate the statistical method employed. The book is extensively referenced, contains many useful exercises for the reader to follow, and has a set of databases that can be used in following these exercises. The book was written to teach and provide reference on statistical methods for research workers in the medical sciences. It assumes little basic knowledge of mathematics and statistics, but requires an intelligent, alert and motivated student with ability in arithmetic. The only omissions that I have noted is a chapter dealing with adaptive clinical trials, which are now being increasingly adopted, and any mention of conjoint analysis, which is extensively used in social sciences and again being increasingly used in quasiobservational studies in medicine. As a primer in its chosen subject, the book has much to recommend it.

Statistics in Medicine by Robert H Riffenburgh is published by Academic Press. ISBN 978-0-12-384864-2

Read full article from PDF >>

Rate this article You must be a member of the site to make a vote.  
Average rating:
0
     

There are no comments in regards to this article.

spacer

Print this page
Send to a friend
Privacy statement
News and Press Releases

AstroNova introduces its range of grass paper labels derived from grass for high quality organic products

Dietzenbach, 29 April 2019, AstroNova GmbH, European headquarters of AstroNova Inc. (RI, USA), responsible for Sales and Customer Service in Europe, the Middle East and Africa (EMEA), is extending its Product Identification range with new approved labelling for direct foodstuff contact.
More info >>

White Papers

The Rare Disease Patients Experience

When planning a clinical trial for a rare disease, it is important to have a holistic understanding of your patient population. You want to be aware of where your patients are coming from, and what it is like to walk in their shoes. These patients need your trial just as much as you need them; even from the earliest stages, conceptualize your rare disease trial as an altruistic relationship between the patient population and your study.
More info >>

 
Industry Events

PDA Europe Annual Meeting 2019

24 June 2019, Hilton Amsterdam

Featuring updates from international regulatory agencies as well as industry, this promises to become another highlight in the 2019 event calendar and is a meeting not to be missed!
More info >>

 

 

©2000-2011 Samedan Ltd.
Add to favourites

Print this page

Send to a friend
Privacy statement