Gerard E. Dallal, Ph.D

Chief, Biostatistics Unit
Jean Mayer USDA Human Nutrition Research Center on Aging
at Tufts University
711 Washington Street
Boston, MA 02111
Gerard.Dallal@tufts.edu

• *Permissions*
• How to cite these pages
• Introductory remarks and an advisory
• Is statistics hard?
• An all-important foundation!
  • Cause & Effect
  • Study design
  • Intention-To-Treat Analysis
  • Meta Analysis
  • Random Samples / Randomization
  • Randomization plans
  • Units of analysis
  • Creating data files
• The basics
  • Look At the Data!
  • Logarithms
  • Summary Statistics
    • Location and Spread
    • Correlation Coefficients
  • Probability
  • The Normal Distribution
  • Outliers
  • The Behavior of Sample Means
    (or Why Confidence Intervals Always Seem to be Based On the Normal Distribution)
• Confidence Intervals and Tests of Signficance
  • Confidence Intervals
  Probability & Statistics / A Boy & His Dog
  CIs for Logarithmically Transformed Data
  LARGE SAMPLE Formulas for Confidence Intervals Involving Population Means
  • Other Intervals
  • Paired Data / Paired Analyses
  What does pairing really do?
  • The Ubiquitous Sample Mean!
  • What Student Did
  • What Student Really Did
  • Significance Tests
  • Prologue
  • Significance Tests/Hypothesis Testing
  • Significance Tests Simplified
  • Student's t Test for Independent Samples
  • P values
  • Why P=0.05?
  • A Valuable Lesson
    • A very nice take on it
    • Exposures
    • Outcomes
    • Oh, what the heck!
  • One-Sided Tests
  • Contingency Tables
  • Proportions
  • Odds
  • Paired Counts
• Sample Size Calculations
• Some Underlying Theory & Some Practical Advice
• An Underappreciated Consequence of Sample Size Calculations As They Are Usually Performed
• Controlled Trials
• Surveys
• Group Randomized, Multi-level, and Hierarchical Studies
• Nonparametric Statistics
• Simple Linear Regression
• Introduction to Simple Linear Regression
• How to Read the Output From Simple Linear Regression Analyses
• Correlation and Regression
• Frank Anscombe's Regression Examples
• Transformations In Linear Regression
• Which fit is better?
• The Regression Effect / The Regression Fallacy
• Comparing Two Measurement Devices: Part I
• Comparing Two Measurement Devices: Part II
• Linear models: Nomenclature
• Multiple Linear Regression
• Introduction to Regression Models
• Student's t Test for Independent Samples Is A Special Case of Simple Linear Regression
• Introduction to Multiple Linear Regression
• The Most Important Lesson You'll Ever Learn About Multiple Linear Regression Analyses
• How to Read the Output From Multiple Linear Regression Analyses
• The Meaning of Regression Coefficents
• What Does Multiple Regression Look Like?
• What Does Multiple Regression Look Like? (Part 2)
• Why Is a Regression Line Straight?
• Partial Correlation Coefficients
• Which Predictors Are More Important?
• The Extra Sum of Squares Principle
• Simplifying A Multiple Regression Equation
• Using the Bootstrap to Simplify a Multiple Regression Equation
• The Real Problem!
• Which variables go into a multiple regression equation?
• The Mechanics of Categorical Variables With More Than Two Categories
• Interactions In Multiple Regression Models
• Regression Diagnostics
  • Collinearity
  Centering
  • Other Regression Diagnostics
• Analysis of Variance
  • Single Factor ANOVA
  • How to Read the Output From One-Way Analysis of Variance
  • Multiple Comparisons
  • Labeling Similar Means After Performing an Analysis of Variance
  • A web page that creates the labels
  • Adjusting Results for Other Variables
  • Adjusted Means, a.k.a. Least Squares Means
  • Adjusted Means: Adjusting For Numerical Variables
  • Adjusted Means: Adjusting For Categorical Variables
  • Which Variables Should We Adjust For?
  • Multi-Factor Analysis of Variance
  • The Model For Two-Factor Analysis of Variance
  • Pooling Effects
  • Fixed and Random Factors
  Randomized (Complete) Block Designs
  • Repeated measures analysis of variance
  • Repeated measures analysis of variance: Part I
  • Repeated measures analysis of variance: Part II
  • Why SAS's PROC MIXED Can Seem So Confusing
  • Pre-test/Post-test Experiments
  • Serial Measurements
• Crossover Studies

A good case can be made that the best set of articles about statistical practice written for the practitioner is the series of Statistics Notes appearing in the British Medical Journal.

There have been many attempts at online statistics instruction. HyperStat is one of the better ones, not only for the content but also for the additional links.

[back to home page]