Health Science Statistics using R and R Commander has been written for students, researchers and professionals who need a practical guide to the subject. R is an open source statistical package that is finding favour in a wide variety of statistics applications. Initially R was the preserve of trained statisticians. However, it is increasingly being used with non-specialist audiences, at both postgraduate and senior undergraduate levels. The book focuses on the graphical user interface, R Commander, which helps make R more user-friendly for the uninitiated. However, throughout the book, the R code behind R Commander is provided to allow the reader to program directly if required. The book provides both the practical skills and essential knowledge to enable the reader to perform their own statistical analyses in R and to interpret the results appropriately. The book starts with introductory chapters which demonstrate how to install and run R and R Commander effortlessly. It then builds from introductory statistics chapters (calculating correlations and t tests) through to more complex areas (structural equation modelling, log linear regression etc.). Each chapter begins with a thorough introduction to the statistical technique under discussion. Then, working through real-life data, the reader is shown how to do their own analysis using R Commander, followed by a demonstration of how to do this analysis in R directly. The later chapters also show how to write up findings in the correct format. For specific analyses other free applications are introduced to supplement R (OpenEpi, Gpower and nyx). Throughout, the reader is given essential tips and advice to help get to grips with carrying out the analysis and intelligently reflecting on the output. Health Science Statistics using R and R Commander is accompanied by an array of web-based material including: additional online chapters discussion board R code for each chapter multiple choice questions links to other resources including websites, blogs and tutorials Health Science Statistics using R and R Commander is a comprehensive introduction to statistics in the health sciences combined with a hands-on practical guide to R (and related free software).
1. How this book works 2. Statistics and R - Setting the scene 3. R - What is it? Two ways to use it 4. Downloading and installing the R software - free! 5. Starting R 6. R Commander: a graphical front end to R 7. Packages: the apps 8. A quick tutorial - Analysing data shipped with R 9. A quick introduction to the R language: R 10. Basic statistical techniques 11. Summary statistics 12. Graphing Distributions of single variables: histograms and density plots 13. Histograms and density plots for subgroups defined by factor levels 14. Boxplots 15. Percentages for each category/factor level 16. Samples and populations 17. Comparing a sample mean to a population mean: Single sample t test 18. Comparing pre-post test means: Paired samples t test 19. Comparing 2 sample means: independent samples t test 20. Comparing pre-post test median difference: Wilcoxon Matched Pairs Statistic 21. Comparing 2 distributions: Mann-Whitney U Statistic 22. Comparing an observed proportion to a population value: The Binomial test 23. Several independent proportions compared with the average: Two way tables 24. Comparing several independent categories: Contingency tables 25. Measuring the degree to which two variable co-vary: Correlation 26. Measuring the influence of one variable on another: Regression 27. Health Statistics 28. Risk and odds ratios 29. Number needed to treat/harm (NNT/NNH) 30. Sensitivity, Specificity, predictive values and likelihood ratios 31. Levels of agreement: Kappa, Krippendorff and the ICC 32. Bland-Altman plots 33. Meta-analysis: the basics 34. Plotting survival over time: K-M (Kaplan-Meier) plots 35. Investigating effects upon survival over time: Cox PH regression 36. Graphical summaries of data: Aggregation 37. Paired nominal data: comparing proportions using McNemar's test 38. Managing your data and R 39. Creating datasets and distributions in R Commander and R 40. Importing your data into R 41. Cutting and Pasting from Excel/Word to the R Data editor 42. Saving and exporting your work and data 43. R Script files (.R) 44. Manipulating variables (columns) in R Commander and R 45. Manipulating cases (rows) in R Commander and R 46. Expanding tables of counts into flat files 47. Installing non-CRANS packages 48. Workspaces, objects and history files 49. Developing R Code: Rstudio and NppToR 50. More ways of analysing your data 51. Mosaic and extended association plots 52. Multiway tables and Crosstabs 53. Resampling: Permutations, Jackknives and Bootstraps 54. Repeated measures: Mixed models and Gee 55. Sample size requirements 56. Confidence intervals for effect sizes: Noncentral distributions 57. Publication quality graphics 58. More Regression Techniques 59. Multiple Linear Regression: Measuring the influence of several variables on a continuous variable 60. Logistic regression: a binary outcome 61. Poisson (log-linear) Regression 62. Conditional Logistic Regression 63. Factorial Anova 64. Factor Analysis 65. Structural Equation Modelling (SEM) 66. Summary Appendices Glossary Index
