Statistics at square one / / edited by Michael J. Campbell

• Pubblicazione: Hoboken, New Jersey ; ; Chichester : , : Wiley Blackwell, , [2021]; ©2021
• Edizione: Twelfth edition.
• Descrizione fisica: 1 online resource (303 pages)
• Disciplina: 610.727
• Soggetto topico: Medical statistics
• Persona (resp. second.): CampbellMichael J. <1950->
• Nota di bibliografia: Includes bibliographical references and index.
• Nota di contenuto: Cover -- Title Page -- Copyright Page -- Contents -- Preface -- About the companion website -- Chapter 1 Understanding basic numbers -- When is a number large? -- Ratios -- Using ratios to adjust for other variables -- Proportions, percentages and odds -- Percentage difference and percentage change: importance of baseline -- Rounding proportions and percentages -- Probabilities and risks -- Prevalence and incidence rate -- Trusting numbers -- Conclusions -- Further reading -- Exercises -- References -- Chapter 2 Data display and summary -- Types of data -- Stem-and-leaf plots and dot plots -- Box-whisker plots -- Median -- Measures of variation -- Frequency tables and histograms -- Bar charts -- Further reading -- Common questions -- What is the distinction between a histogram and a bar chart? -- What are poor methods of displaying data? -- Displaying data in papers -- Exercises -- References -- Chapter 3 Summary statistics for quantitative data -- Mean -- Variance and standard deviation -- Standard deviation from ungrouped data -- Standard deviation from grouped data -- Normal distribution -- Skewness -- Between-subjects and within-subjects standard deviation -- Common questions -- When should I quote the mean and when should I quote the median to describe my data? -- When should I use a standard deviation to summarise variability? -- How can I tell if data are skewed from a table? -- When should I use the mode? -- Formula appreciation -- Reading and Displaying Summary Statistics -- Exercises -- References -- Chapter 4 Summary statistics for binary data -- Summarising one binary variable -- Summarising the relationship between two binary variables -- Relative Risks versus Odds Ratios -- Odds ratios and cross-sectional studies -- Odds ratios and case-control studies -- Example of a case-control study. ; Estimating relative risk from case-control studies -- Common questions -- When should I quote an odds ratio and when should I quote a relative risk? -- How does one choose the numerator and denominator for a relative risk? -- How should one quote relative risks? -- Should one ever quote a number needed to treat? -- Reading and displaying summary statistics -- Exercises -- References -- Chapter 5 Diagnostic and screening tests -- Diagnostic and screening tests -- Examples -- Example 1: Test for COVID-19 -- Example 2: Test for generalised anxiety disorder -- Sensitivity and Specificity -- Positive predictive value in relation to prevalence -- Likelihood ratio -- Receiver operating characteristics curves -- Further discussion on diagnostic and screening tests -- Limitations of the conventional diagnostic testing paradigm -- Reading and reporting diagnostic/screening tests -- Exercises -- References -- Chapter 6 Populations and samples -- Populations -- Samples -- Unbiasedness and precision -- Problems of bias in non-randomised samples (especially Big Data) -- Randomisation -- Variation between samples -- Standard error of the mean -- Example of standard error -- Standard error of a proportion or a percentage -- Problems with non-random samples -- Common questions -- What is an acceptable response rate from a survey? -- Given measurements on a sample, what is the difference between a standard deviation and a standard error? -- When should I use a standard deviation to describe data and when should I use a standard error? -- Important points -- Reading and reporting populations and samples -- Exercises -- References -- Chapter 7 Statements of probability and confidence intervals -- Reference ranges -- Confidence intervals -- Large sample standard error of difference between means -- Large sample confidence interval for the difference in two means. ; Standard error of difference between percentages or proportions -- Confidence interval for a difference in proportions or percentages -- Confidence interval for an odds ratio -- Confidence interval for a relative risk -- Confidence Intervals for other estimates -- Common Questions -- What is the difference between a reference range and a confidence interval? -- If I repeated a study with the same sample size, would the new results fall in the confidence interval 95% of the time? -- Reading and reporting confidence intervals -- Formula appreciation -- Exercises -- References -- Chapter 8 P values, power, type I and type II errors -- Null hypothesis and type I error -- Testing for differences of two means -- Testing for a difference in two proportions -- P value -- P values, confidence intervals and clinically important results -- Alternative hypothesis and type II error -- Other types of statistical inference -- Issues with P values -- One-sided and two-sided tests -- Tests for superiority, tests for non-inferiority and tests for equivalence -- Links with diagnostic tests -- Common questions -- Why is the P value not the probability that the null hypothesis is true? -- Why is 5% usually used as the level at which results are deemed 'significant'? -- Reading and reporting significance tests -- Exercises -- References -- Chapter 9 Tests for differences between two groups of a quantitative outcome with small samples -- Student's t test -- Confidence interval for the mean from a small sample -- Difference of sample mean from population mean (one-sample t test) -- Difference between means of two samples -- Unequal standard deviations -- Difference between means of paired samples (paired t test) -- Non-parametric or distribution-free tests -- Tests for differences in unpaired samples of non-Normally distributed data (Mann-Whitney U test). ; Tests for differences in paired samples of non-Normally distributed data (Wilcoxon test) -- Computer-intensive methods -- Permutation tests: unpaired tests -- Permutation tests: paired tests -- The bootstrap -- Discussion -- Reading and reporting t tests and non-parametric tests -- Common questions -- Should I test my data for Normality before using the t test? -- Should I test for equality of the standard deviations before using the usual t test? -- Why should I use a paired test if my data are paired? What happens if I don't? -- Do non-parametric tests compare medians? -- How is the Mann-Whitney U test related to the t test? -- How is the Mann-Whitney U test related to the area under the receiver operating characteristics curve of Chapter 5? -- References -- Chapter 10 Tests for association in binary and categorical data -- General chi-squared test -- 2 × 2 tables -- Small numbers: Yates' correction, Fisher's Exact Test and the permutation test -- 2 test for trend -- Comparison of an observed and a theoretical distribution -- Tests for paired binary data -- Examples of a paired comparison -- Extensions of the 2 test -- Common questions -- There are a number of tests of association for a 2 × 2 table. Which should I choose? -- I have matched data, but the matching criteria were very weak. Should I use McNemar's test? -- Do chi-squared tests apply to large contingency tables? -- Is the chi-squared test a non-parametric test? -- Formula appreciation -- Reading and reporting chi-squared tests -- Exercises -- References -- Chapter 11 Correlation and regression -- The correlation coefficient -- Looking at data: scatter diagrams -- Calculation of the correlation coefficient -- Significance test for a correlation coefficient -- Spearman rank correlation -- The regression equation -- Simple checks of the model -- Using regression in t tests. ; More advanced methods -- Common questions -- If two variables are correlated, are they causally related? -- How do I test the assumptions underlying linear regression? -- When should I use correlation and when should I use regression? -- Which are the important assumptions for linear regression? -- Formula appreciation -- Reading and reporting correlation and regression -- Exercises -- References -- Chapter 12 Survival analysis -- Why survival analysis is different -- Kaplan-Meier survival curve -- Example of calculation of survival curve -- The log rank test -- Further methods -- Common questions -- Do I need to test for a constant relative risk before doing the log rank test? -- If I don't have any censored observations, do I need to use survival analysis? -- How does the hazard calculated under the log rank compare with the usual estimate of risk? -- Reading and reporting survival analysis -- Exercises -- References -- Chapter 13 Modelling data -- Basics -- Models -- Model fitting and analysis: exploratory and confirmatory analyses -- Bayesian methods -- Models generally -- X1 binary and X2 binary -- X1 continuous and X2 continuous -- X1 binary and X2 continuous -- Multiple linear regression -- Example linear regression -- Paper critique -- Logistic regression -- Logistic regression instead of a chi-squared test -- Example of logistic regression from the literature -- Paper critique -- Survival analysis -- Proportional hazards models -- Proportional hazards model instead of log rank -- Example of proportional hazards model -- Paper critique -- Other things to consider in modelling -- References -- Chapter 14 Study design and choosing a statistical test -- Design -- Sample size -- Choice of test -- Reading and reporting on the design of a study -- Further reading -- Exercises -- References -- Chapter 15 Use of computer software. ; Chapter 2: Data display and summary.
• Sommario/riassunto: "This book is aimed at anyone who needs a basic introduction to statistics in the health sciences. It is based on many years' experience teaching first year medical and health science students. Many of the examples are taken from Primary Care in the UK, which is where I worked for many years. Throughout I have tried to emphasise that Medical Statistics is not just a bag of tricks, and there are many synergies between the methods. It is now over forty years since Swinscow's original edition, and each edition reflected changes in the understanding of medical statistics. Perhaps the greatest change has occurred since the previous edition,which appeared twelve years ago.Despite the efforts of medical statisticians, there was a widespread misuse of p-values,the cornerstone of conventional statistical inference. This led some journals to ban p-values altogether. It is my view that used properly the p-value is a useful concept but this book, as in previous editions of this book, concentrates on estimation rather than just hypothesis testing. The book tries to steer the reader away from an excessive devotion to p-values, to instil a proper appreciation of their usefulness and to emphasise estimation over significance testing"--
• Titolo autorizzato: Statistics at square one
• ISBN: 9781119402350; 1-119-40234-4; 1-119-40235-2; 1-119-40142-9
• Formato: Materiale a stampa
• Livello bibliografico: Monografia
• Lingua di pubblicazione: Inglese
• Record Nr.: 9910830466503321
• Lo trovi qui: Univ. Federico II