Details

Preface . <p>Statistical packages.</p> <p><b>PART 1. DATA TYPES.</b></p> <p><b>1. Data types.</b></p> <p>1.1Does it really matter?</p> <p>1.2 Interval scale data.</p> <p>1.3 Ordinal scale data.</p> <p>1.4 Nominal scale data.</p> <p>1.5 Structure of this book.</p> <p>1.6 Chapter summary.</p> <p><b>PART 2. INTERVAL-SCALE DATA.</b></p> <p><b>2. Descriptive statistics.</b></p> <p>2.1 Summarizing data sets.</p> <p>2.2 Indicators of central tendency. mean, median and mode.</p> <p>2.3 Describing variability. standard deviation and coefficient of variation.</p> <p>2.4 Quartiles. another way to describe data .</p> <p>2.5 Using computer packages to generate descriptive statistics.</p> <p>2.6 Chapter summary.</p> <p><b>3. The normal distribution.</b></p> <p>3.1 What is a normal distribution? .</p> <p>3.2 Identifying data that are not normally distributed.</p> <p>3.3 Proportions of individuals within one or two standard deviations of the mean.</p> <p>3.4 Chapter summary.</p> <p><b>4. Sampling from populations. the SEM.</b></p> <p>4.1 Samples and populations.</p> <p>4.2 From sample to population.</p> <p>4.3 Types of sampling error.</p> <p>4.4 What factors control the extent of random sampling error?</p> <p>4.5 Estimating likely sampling error. The SEM.</p> <p>4.6 Offsetting sample size against standard deviation.</p> <p>4.7Chapter summary.</p> <p><b>5. Ninety-five per cent confidence interval for the mean.</b></p> <p>5.1 What is a confidence interval?</p> <p>5.2 How wide should the interval be?</p> <p>5.3 What do we mean by ‘95 per cent’ confidence?</p> <p>5.4 Calculating the interval width.</p> <p>5.5 A long series of samples and 95 per cent confidence intervals.</p> <p>5.6 How sensitive is the width of the confidence interval to changes in the SD, the sample size or the required level of confidence?</p> <p>5.7 Two statements.</p> <p>5.8 One-sided 95 per cent confidence intervals.</p> <p>5.9 The 95 per cent confidence interval for the difference between two treatments.</p> <p>5.10 The need for data to follow a normal distribution and data transformation.</p> <p>5.11 Chapter summary.</p> <p><b>6. The two-sample t-test(1).Introducing hypothesis tests.</b></p> <p>6.1 The two-sample t-test. an example of a hypothesis test.</p> <p>6.2 ‘Significance’.</p> <p>6.3 The risk of a false positive finding.</p> <p>6.4 What factors will influence whether or not we obtain a significant outcome?</p> <p>6.5 Requirements for applying a two-sample t-test.</p> <p>6.6 Chapter summary.</p> <p><b>7. The two-sample t-test(2).The dreaded P value.</b></p> <p>7.1 Measuring how significant a result is.</p> <p>7.2 P values.</p> <p>7.3 Two ways to define significance?</p> <p>7.4 Obtaining the P value.</p> <p>7.5 P values or 95 per cent confidence intervals?</p> <p>7.6 Chapter summary.</p> <p><b>8. The two-sample t-test(3).False negatives, power and necessary sample sizes.</b></p> <p>8.1 What else could possibly go wrong?</p> <p>8.2 Power.</p> <p>8.3 Calculating necessary sample size.</p> <p>8.4 Chapter summary.</p> <p><b>9. The two-sample t-test(4).Statistical significance, practical significance and equivalence.</b></p> <p>9.1 Practical significance. is the difference big enough to matter?</p> <p>9.2 Equivalence testing.</p> <p>9.3 Non-inferiority testing.</p> <p>9.4 P values are less informative and can be positively misleading.</p> <p>9.5 Setting equivalence limits prior to experimentation.</p> <p>9.6 Chapter summary.</p> <p><b>10. The two-sample t-test(5).One-sided testing.</b></p> <p>10.1 Looking for a change in a specified direction.</p> <p>10.2 Protection against false positives.</p> <p>10.3 Temptation!.</p> <p>10.4 Using a computer package to carry out a one-sided test.</p> <p>10.5 Should one-sided tests be used more commonly?</p> <p>10.5 Chapter summary.</p> <p><b>11. What does a statistically significant result really tell us?</b></p> <p>11.1 Interpreting statistical significance.</p> <p>11.2 Starting from extreme scepticism.</p> <p>11.3 Chapter summary.</p> <p><b>12. The paired t-test. comparing two related sets of measurements.</b></p> <p>12.1 Paired data.</p> <p>12.2 We could analyse the data using a two-sample t-test.</p> <p>12.3 Using a paired t-test instead.</p> <p>12.4 Performing a paired t-test.</p> <p>12.5 What determines whether a paired t-test will be significant?</p> <p>12.6 Greater power of a paired t-test.</p> <p>12.7 The paired t-test is only applicable to naturally paired data.</p> <p>12.8 Choice of experimental design.</p> <p>12.9 Requirements for applying a paired t-test .</p> <p>12.10 Sample sizes, practical significance and one-sided tests.</p> <p>12.11 Summarizing the differences between the paired and two-sample t-tests.</p> <p>12.12 Chapter summary.</p> <p><b>13. Analyses of variance. going beyond t-tests.</b></p> <p>13.1 Extending the complexity of experimental designs.</p> <p>13.2 One-way analysis of variance.</p> <p>13.3 Two-way analysis of variance.</p> <p>13.4 Multi-factorial experiments.</p> <p>13.5 Keep it simple. Keep it powerful.</p> <p>13.6 Chapter summary.</p> <p><b>14. Correlation and regression. relationships between measured values.</b></p> <p>14.1 Correlation analysis.</p> <p>14.2 Regression analysis.</p> <p>14.3 Multiple regression.</p> <p>14.4 Chapter summary.</p> <p><b>PART 3. NOMINAL-SCALE DATA.</b></p> <p><b>15. Describing categorized data.</b></p> <p>15.1 Descriptive statistics.</p> <p>15.2 Testing whether the population proportion might credibly be some pre-determined figure.</p> <p>15.3 Chapter summary.</p> <p><b>16. Comparing observed proportions. the contingency chi-square test.</b></p> <p>16.1 Using the contingency chi-square test to compare observed proportions.</p> <p>16.2 Obtaining a 95 per cent CI for the change in the proportion of expulsions. is the difference large enough to be of practical significance?</p> <p>16.3 Larger tables. attendance at diabetic clinics.</p> <p>16.4 Planning experimental size.</p> <p>16.5 Chapter summary.</p> <p><b>PART 4. ORDINAL-SCALE DATA.</b></p> <p><b>17. Ordinal and non-normally distributed data.Transformations and non-parametric tests.</b></p> <p>17.1 Transforming data to a normal distribution.</p> <p>17.2 The Mann-Whitney test. a non-parametric method.</p> <p>17.3 Dealing with ordinal data.</p> <p>17.4 Other non-parametric methods.</p> <p>17.5 Chapter summary.</p> <p>Appendix to Chapter 17.</p> <p><b>PART 5. SOME CHALLENGES FROM THE REAL WORLD.</b></p> <p><b>18. Multiple testing.</b></p> <p>18.1 What is it and why is it a problem?</p> <p>18.2 Where does multiple testing arise?</p> <p>18.3 Methods to avoid false positives.</p> <p>18.4 The role of scientific journals.</p> <p>18.5 Chapter summary.</p> <p><b>19. Questionnaires.</b></p> <p>19.1 Is there anything special about questionnaires?</p> <p>19.2 Types of questions.</p> <p>19.3 Designing a questionnaire.</p> <p>19.4 Sample sizes and return rates.</p> <p>19.5 Analysing the results.</p> <p>19.6 Confounded epidemiological data.</p> <p>19.7 Multiple testing with questionnaire data.</p> <p>19.8 Chapter summary.</p> <p><b>PART 6. CONCLUSIONS.</b></p> <p><b>20. Conclusions.</b></p> <p>20.1 Be clear about the purpose of the experiment.</p> <p>20.2 Keep the experimental design simple and therefore clear and powerful.</p> <p>20.3 Draw up a statistical analysis plan as part of the experimental design. it is not a last minute add-on.</p> <p>20.4 Explore your data visually before launching into statistical testing.</p> <p>20.5 Beware of multiple analyses.</p> <p>20.6 Interpret both significance and non-significance with care.</p> <p>Index.</p>

<p><strong>Dr Philip Rowe.</strong> Reader in Pharmaceutical Computing, School of Pharmacy and Chemistry, Liverpool,?UK<br />In addition to Dr Rowe's teaching and research at LJMU he also works on a consultancy basis offering advice and assistance with pharmacokinetic or general data analysis problems for the pharmaceutical industry, professional organizations and hospitals. He has recently secured a post delivering statistics training for the Institute of Clinical Research.

GB