Location of Tables for Tests of Significance 

xv  
Preface 

xvii  

Biostatistics and Clinical Practice 


1  (10) 


2  (3) 

What Do Statistical Procedures Tell You? 


5  (2) 

Why Not Depend on the Journals? 


7  (2) 

Why Has the Problem Persisted? 


9  (2) 


11  (29) 


13  (1) 


14  (1) 


15  (1) 


16  (5) 


21  (9) 


21  (3) 


24  (2) 

Experiments and Observational Studies 


26  (3) 

Randomized Clinical Trials 


29  (1) 

How to Estimate the Mean and Standard Deviation from a Sample 


30  (1) 

How Good Are These Estimates? 


31  (7) 


38  (1) 


38  (2) 

How to Test for Differences between Groups 


40  (33) 


41  (5) 

Two Different Estimates of the Population Variance 


46  (2) 


48  (8) 


56  (11) 

Glucose Levels in Children of Parents with Diabetes 


56  (4) 

Halothane versus Morphine for OpenHeart Surgery 


60  (4) 

Menstrual Dysfunction in Distance Runners 


64  (3) 


67  (6) 

The Special Case of Two Groups: The t Test 


73  (53) 


75  (2) 

The Standard Deviation of a Difference or a Sum 


77  (3) 

Use of t to Test Hypotheses about Two Groups 


80  (7) 

What If the Two Samples Are Not the Same Size? 


87  (1) 


88  (4) 

Glucose Levels in Children of Parents with Diabetes 


88  (1) 

Halothane versus Morphine for OpenHeart Surgery 


88  (4) 

The t Test Is an Analysis of Variance 


92  (3) 

Common Errors in the Use of the t Test and How to Compensate for Them 


95  (3) 

How to Use t Tests to Isolate Differences between Groups in Analysis of Variance 


98  (8) 


98  (2) 

More on Menstruation and Jogging 


100  (1) 

A Better Approach to Multiple Comparisons: The Holm t Test 


101  (3) 


104  (2) 

Other Approaches to Multiple Comparison Testing: The StudentNewmanKeuls Test 


106  (5) 

Still More on Menstruation and Jogging 


107  (3) 


110  (1) 

Which Multiple Comparison Procedure Should You Use? 


111  (1) 

Multiple Comparisons against a Single Control 


112  (5) 


112  (1) 


113  (1) 


113  (4) 


117  (6) 

Statistical versus Real (Clinical) Thinking 


119  (2) 


121  (2) 


123  (3) 

How to Analyze Rates and Proportions 


126  (53) 


127  (5) 

Estimating Proportions from Samples 


132  (5) 

Hypothesis Tests for Proportions 


137  (8) 

The Yates Correction for Continuity 


139  (1) 

Mortality Associated with Anesthesia for OpenHeart Surgery with Halothane or Morphine 


140  (2) 

Prevention of Thrombosis in People Receiving Hemodialysis 


142  (3) 

Another Approach to Testing Nominal Data: Analysis of Contingency Tables 


145  (7) 

The ChiSquare Test Statistic 


148  (4) 

ChiSquare Applications to Experiments with More Than Two Treatments or Outcomes 


152  (6) 

Subdividing Contingency Tables 


155  (3) 


158  (5) 

Measures of Association Between Two Nominal Variables 


163  (7) 

Prospective Studies and Relative Risk 


164  (2) 

CaseControl Studies and the Odds Ratio 


166  (2) 

Passive Smoking and Breast Cancer 


168  (2) 


170  (9) 

What Does ``Not Significant'' Really Mean? 


179  (40) 


180  (5) 


185  (1) 

What Determines a Test's Power? 


186  (18) 

The Size of the Type I Error α 


187  (6) 

The Size of the Treatment Effect 


193  (2) 

The Population Variability 


195  (2) 

Bigger Samples Mean More Powerful Tests 


197  (4) 

What Determines Power? A Summary 


201  (2) 

Another Look at Halothane versus Morphine for OpenHeart Surgery 


203  (1) 

Power and Sample Size for Analysis of Variance 


204  (4) 

Power, Menstruation, and Running 


206  (2) 

Power and Sample Size for Comparing Two Proportions 


208  (3) 

Mortality Associated with Anesthesia for OpenHeart Surgery 


210  (1) 

Sample Size for Comparing Two Proportions 


211  (1) 

Power and Sample Size for Relative Risk and Odds Ratio 


211  (1) 

Power and Sample Size for Contingency Tables 


212  (2) 

Physicians, Perspiration, and Power 


213  (1) 

Practical Problems in Using Power 


214  (1) 

What Difference Does It Make? 


215  (3) 


218  (1) 


219  (34) 

The Size of the Treatment Effect Measured as the Difference of Two Means 


220  (3) 


223  (4) 


224  (3) 

What Does ``Confidence'' Mean? 


227  (2) 

Confidence Intervals Can Be Used to Test Hypotheses 


229  (2) 

Confidence Interval for the Population Mean 


231  (2) 

The Size of the Treatment Effect Measured as the Difference of Two Rates or Proportions 


233  (7) 

Difference in Mortality Associated with Anesthesia for OpenHeart Surgery 


234  (1) 

Difference in Thrombosis with Aspirin in People Receiving Hemodialysis 


235  (1) 

How Negative Is a ``Negative'' Clinical Trial? 


236  (1) 


236  (4) 

Confidence Interval for Rates and Proportions 


240  (5) 

Quality of Evidence Used as a Basis for Interventions to Improve Hospital Antibiotic Prescribing 


241  (1) 

Exact Confidence Intervals for Rates and Proportions 


242  (3) 

Confidence Intervals for Relative Risk and Odds Ratio 


245  (2) 

Difference in Thrombosis with Aspirin in People Receiving Hemodialysis 


246  (1) 

Passive Smoking and Breast Cancer 


247  (1) 

Confidence Interval for the Entire Population 


247  (4) 


251  (2) 


253  (68) 


254  (5) 

The Population Parameters 


257  (2) 

How to Estimate the Trend from a Sample 


259  (19) 

The Best Straight Line through the Data 


259  (8) 

Variability about the Regression Line 


267  (1) 

Standard Errors of the Regression Coefficients 


268  (4) 

How Convincing Is the Trend? 


272  (2) 

Confidence Interval for the Line of Means 


274  (1) 

Confidence Interval for an Observation 


275  (3) 

How to Compare Two Regression Lines 


278  (7) 

Overall Test for Coincidence of Two Regression Lines 


280  (1) 

Relationship between Weakness and Muscle Wasting in Rheumatoid Arthritis 


281  (4) 

Correlation and Correlation Coefficients 


285  (11) 

The Pearson ProductMoment Correlation Coefficient 


287  (3) 

The Relationship between Regression and Correlation 


290  (2) 

How to Test Hypotheses about Correlation Coefficients 


292  (1) 

Journal Size and Selectivity 


293  (3) 

The Spearman Rank Correlation Coefficient 


296  (6) 

Variation among Interns in Use of Laboratory Tests: Relation to Quality of Care 


300  (2) 

Power and Sample Size in Regression and Correlation 


302  (3) 

Comparing Two Different Measurements of the Same Thing: The BlandAltman Method 


305  (5) 

Assessing Mitral Regurgitation with Echocardiography 


306  (4) 


310  (1) 


310  (11) 

Experiments When Each Subject Receives More than One Treatment 


321  (42) 

Experiments When Subjects Are Observed before and after a Single Treatment: The Paired t Test 


322  (8) 

Cigarette Smoking and Platelet Function 


325  (5) 

Another Approach to Analysis of Variance 


330  (12) 


331  (8) 

Accounting for All the Variability in the Observations 


339  (3) 

Experiments When Subjects Are Observed after Many Treatments: RepeatedMeasures Analysis of Variance 


342  (12) 

AntiAsthmatic Drugs and Endotoxin 


348  (4) 

How to Isolate Differences in RepeatedMeasures Analysis of Variance 


352  (1) 

Power in RepeatedMeasures Analysis of Variance 


353  (1) 

Experiments When Outcomes Are Measured on a Nominal Scale: McNemar's Test 


354  (3) 

p7 Antigen Expression in Human Breast Cancer 


354  (3) 


357  (6) 

Alternatives to Analysis of Variance and the t Test Based on Ranks 


363  (50) 

How to Choose between Parametric and Nonparametric Methods 


364  (3) 

Two Different Samples: The MannWhitney RankSum Test 


367  (11) 

The Leboyer Approach to Childbirth 


374  (4) 

Each Subject Observed before and after One Treatment: The Wilcoxon SignedRank Test 


378  (8) 

Cigarette Smoking and Platelet Function 


385  (1) 

Experiments with Three or More Groups When Each Group Contains Different Individuals: The KruskalWallis Statistic 


386  (9) 

Prenatal Marijuana Exposure and Child Behavior 


389  (1) 

Nonparametric Multiple Comparisons 


390  (2) 


392  (3) 

Experiments in Which Each Subject Receives More than One Treatment: The Friedman Test 


395  (10) 

AntiAsthmatic Drugs and Endotoxin 


399  (2) 

Multiple Comparisons after Friedman's Test 


401  (1) 

Effect of Secondhand Smoke on Angina Pectoris 


401  (4) 


405  (1) 


406  (7) 

How to Analyze Survival Data 


413  (31) 


414  (3) 

Estimating the Survival Curve 


417  (10) 


423  (1) 

Standard Errors and Confidence Limits for the Survival Curve 


424  (3) 

Comparing Two Survival Curves 


427  (10) 

Bone Marrow Transplantation to Treat Adult Leukemia 


429  (7) 

The Yates Correction for the LogRank Test 


436  (1) 


437  (1) 


438  (2) 


440  (1) 


440  (4) 

What Do the Data Really Show? 


444  (22) 


445  (2) 


447  (8) 

Internal Mammary Artery Ligation to Treat Angina Pectoris 


448  (1) 

The Portacaval Shunt to Treat Cirrhosis of the Liver 


449  (3) 

Is Randomization of People Ethical? 


452  (2) 

Is a Randomized Controlled Trial Always Necessary? 


454  (1) 

Does Randomization Ensure Correct Conclusions? 


455  (5) 

Problems with the Population 


460  (2) 

How You Can Improve Things 


462  (4) 
Appendix A Computational Forms 

466  (6) 
Appendix B Power Charts 

472  (9) 
Appendix C Answers to Exercises 

481  (16) 
Index 

497  