Statistics at Square Two: Understanding Modern Statistical Applications in Medicine

by Campbell, Michael J.

• ISBN13: 9780727913944
• ISBN10: 0727913948
• Edition: 10th
• Format: Paperback
• Copyright: 2001-08-01
• Publisher: John Wiley & Sons Inc

Summary

Statistical methods used in the medical literature are becoming increasingly complex, but the software available to apply these methods is becoming easier to use. The result is that consumers of the literature now have to grasp sophisticated methods which may be inappropriately applied. This book, which goes beyond the basics of Statistics at Square One, reviews the most commonly used modern statistical methods and highlights common misunderstandings. It is easy to read, with annotated computer outputs and a minimum of formulas.

Table of Contents

Preface	ix
Models, tests and data	1	(11)
Basics	1	(1)
Models	2	(1)
Types of data	3	(2)
Significance tests	5	(1)
Confidence intervals	6	(1)
Statistical tests using models	7	(1)
Model fitting and analysis: exploratory and confirmatory analyses	7	(1)
Computer-intensive methods	8	(1)
Bayesian Methods	8	(1)
Reporting statistical results in the literature	9	(1)
Reading statistics in the literature	10	(2)
Multiple choice questions	10	(2)
Multiple linear regression	12	(25)
The model	12	(1)
Uses of multiple regression	13	(1)
Two independent variables	14	(5)
Interpreting a computer output	19	(6)
Multiple regression in action	25	(1)
Assumptions underlying the models	26	(2)
Model sensitivity	28	(3)
Stepwise regression	31	(1)
Reporting the results of a multiple regression	32	(1)
Reading the results of a multiple regression	32	(5)
Frequently asked questions	33	(2)
Multiple choice questions	35	(2)
Logistic regression	37	(22)
The model	37	(3)
Uses of logistic regression	40	(1)
Interpreting a computer output: grouped analysis	40	(4)
Logistic regression in action	44	(1)
Model checking	45	(2)
Interpreting a computer output: ungrouped analysis	47	(2)
Case---control studies	49	(1)
Interpreting a computer output: unmatched case---control study	50	(1)
Matched case---control studies	51	(1)
Interpreting a computer output: matched case---control study	52	(3)
Conditional logistic regression in action	55	(1)
Reporting the results of logistic regression	55	(1)
Reading about logistic regression	56	(3)
Frequently asked questions	56	(1)
Multiple choice questions	57	(2)
Survival analysis	59	(15)
Introduction	59	(1)
The Model	60	(2)
Uses of Cox regression	62	(1)
Interpreting a computer output	62	(3)
Survival analysis in action	65	(1)
Interpretation of the model	65	(1)
Generalisations of the model	66	(1)
Model checking	67	(2)
Reporting the results of a survival analysis	69	(1)
Reading about the results of a survival analysis	70	(4)
Frequently asked question	70	(1)
Exercise	70	(4)
Random effects models	74	(14)
Introduction	74	(1)
Models for random effects	75	(1)
Random vs fixed effects	76	(1)
Use of random effects models	77	(1)
Random effects models in action	78	(1)
Ordinary least squares at the group level	79	(1)
Computer analysis	80	(4)
Model checking	84	(1)
Reporting the results of a random effects analysis	85	(1)
Reading about the results of a random effects analysis	85	(3)
Frequently asked question	86	(2)
Other models	88	(15)
Poisson regression	88	(4)
Ordinal regression	92	(5)
Time-series regression	97	(3)
Reporting Poisson, ordinal or time-series regression in the literature	100	(1)
Reading about the results of Poisson, ordinal or time-series regression in the literature	101	(2)
Appendix 1 Exponentials and logarithms	103	(2)
Appendix 2 Maximum likelihood and significance tests	105	(10)
Appendix 3 Bootstrapping	115	(6)
Appendix 4 Bayesian Methods	121	(4)
Answers to exercises	125	(2)
Index	127