Principles & Methods of Statistical Analysis

Jerome Frieman earned his PhD from Kent State University in Ohio. He has been at Kansas State University since 1968. Over the course of his career, he engaged in research on operant conditioning in pigeons, rats, and dwarf hamsters; Pavlovian conditioning in rats; social learning in dwarf hamsters; and extraordinary memory in a human participant. He is the author of Learning and Adaptive Behavior and co-author of Memory Search by a Memorist.

Preface About the Authors Prologue PART I * GETTING STARTED Chapter 1: The Big Picture Models The Classical Statistical Model Designing Experiments and Analyzing Data Summary Questions Raised by the Use of the Classical Statistical Model Conceptual Exercises Chapter 2: Examining Our Data: An Introduction to Some of the Techniques of Exploratory Data Analysis Descriptive Statistics Histograms Exploratory Data Analysis Quantile Plots Stem-and-Leaf Displays Letter-Value Displays Box Plots Did My Data Come From a Normal Distribution? Why Should We Care About Looking at Our Data? Summary Conceptual Exercises PART II * THE BEHAVIOR OF DATA Chapter 3: Properties of Distributions: The Building Blocks of Statistical Inference The Effects of Adding a Constant or Multiplying by a Constant The Standard Score Transformation The Effects of Adding or Subtracting Scores From Two Different Distributions The Distribution of Sample Means The Central Limit Theorem Averaging Means and Variances Expected Value Theorems on Expected Value Summary Conceptual Exercises PART III * THE BASICS OF STATISTICAL INFERENCE: DRAWING CONCLUSIONS FROM OUR DATA Chapter 4: Estimating Parameters of Populations From Sample Data Statistical Inference With the Classical Statistical Model Criteria for Selecting Estimators of Population Parameters Maximum Likelihood Estimation Confidence Intervals Beyond Normal Distributions and Estimating Population Means Summary Conceptual Exercises Chapter 5: Resistant Estimators of Parameters A Closer Look at Sampling From Non-Normal Populations The Sample Mean and Sample Median Are L-Estimators Measuring the Influence of Outliers on Estimates of Location and Spread ?-Trimmed Means as Resistant and Efficient Estimators of Location Winsorizing: Another Way to Create a Resistant Estimator of Location Applying These Resistant Estimators to Our Data Resistant Estimators of Spread Applying These Resistant Estimators to Our Data (Part 2) M-Estimators: Another Approach to Finding Resistant Estimators of Location Which Estimator of Location Should I Use? Resampling Methods for Constructing Confidence Intervals A Final Caveat Summary Conceptual Exercises Chapter 6: General Principles of Hypothesis Testing Experimental and Statistical Hypotheses Estimating Parameters The Criterion for Evaluating Our Statistical Hypotheses Creating Our Test Statistic Drawing Conclusions About Our Null Hypothesis But Suppose H0 Is False? Errors in Hypothesis Testing Power and Power Functions The Use of Power Functions p-Values, a, and Alpha (Type I) Errors: What They Do and Do Not Mean A Word of Caution About Attempting to Estimate the Power of a Hypothesis Test After the Data Have Been Collected Is It Ever Appropriate to Use a One-Tailed Hypothesis Test? What Should We Mean When We Say Our Results Are Statistically Significant? A Final Word Summary Conceptual Exercises PART IV * SPECIFIC TECHNIQUES TO ANSWER SPECIFIC QUESTIONS Chapter 7: The Independent Groups t-Tests for Testing for Differences Between Population Means Student's t-test Distribution of the Independent Groups t-Statistic when H0 Is True Distribution of the Independent Groups t-Statistic When H0 Is False Factors That Affect the Power of the Independent Groups t-Test The Assumption Behind the Homogeneity of Variance Assumption Graphical Methods for Comparing Two Groups Suppose the Population Variances Are Not Equal? Standardized Group Differences as Estimators of Effect Size Robust Hypothesis Testing Resistant Estimates of Effect Size Summary Conceptual Exercises Chapter 8: Testing Hypotheses When the Dependent Variable Consists of Frequencies of Scores in Various Categories Classifying Data Testing Hypotheses When the Dependent Variable Consists of Only Two Possibilities The Binomial Distribution Testing Hypotheses About the Parameter p in a Binomial Experiment The Normal Distribution Approximation to the Binomial Distribution Testing Hypotheses About the Difference Between Two Binomial Parameters (p1 - p2) Testing Hypotheses in Which the Dependent Variable Consists of Two or More Categories Summary Conceptual Exercises Chapter 9: The Randomization/Permutation Model: An Alternative to the Classical Statistical Model for Testing Hypotheses About Treatment Effects The Assumptions Underlying the Classical Statistical Model The Assumptions Underlying the Randomization Model Hypotheses for Both Models The Exact Randomization Test for Testing Hypotheses About the Effects of Different Treatments on Behavior The Approximate Randomization Test for Testing Hypotheses About the Effects of Different Treatments on Behavior Using the Randomization Model to Investigate Possible Effects of Treatments Single-Participant Experimental Designs Summary Conceptual Exercises Additional Resources Chapter 10: Exploring the Relationship Between Two Variables: Correlation Measuring the Degree of Relationship Between Two Interval-Scale Variables Randomization (Permutation) Model for Testing Hypotheses About the Relationship Between Two Variables The Bivariate Normal Distribution Model for Testing Hypotheses About Population Correlations Creating a Confidence Interval for the Population Correlation Using the Bivariate Normal Distribution Model Bootstrap Confidence Intervals for the Population Correlation Unbiased Estimators of the Population Correlation Robust Estimators of Correlation Assessing the Relationship Between Two Nominal Variables The Fisher Exact Probability Test for 2 x 2 Contingency Tables With Small Sample Sizes Correlation Coefficients for Nominal Data in Contingency Tables Summary Conceptual Exercises Chapter 11: Exploring the Relationship Between Two Variables: The Linear Regression Model Assumptions for the Linear Regression Model Estimating Parameters With the Linear Regression Model Regression and Prediction Variance and Correlation Testing Hypotheses With the Linear Regression Model Summary Conceptual Exercises Chapter 12: A Closer Look at Linear Regression The Importance of Looking at Our Data Using Residuals to Check Assumptions Testing Whether the Relationship Between Two Variables Is Linear The Correlation Ratio: An Alternate Way to Measure the Degree of Relationship and Test for a Linear Relationship Where Do We Go From Here? When the Relationship Is Not Linear The Effects of Outliers on Regression Robust Alternatives to the Method of Least Squares A Quick Peek at Multiple Regression Summary Conceptual Exercises Chapter 13: Another Way to Scale the Size of Treatment Effects The Point Biserial Correlation Coefficient and the t-Test Advantages and Disadvantages of Estimating Effect Sizes With Correlation Coefficients or Standardized Group Difference Measures Confidence Intervals for Effect Size Estimates Final Comments on the Use of Effect Size Estimators Summary Conceptual Exercises Chapter 14: Analysis of Variance for Testing for Differences Between Population Means What Are the Sources of Variation in Our Experiments? Experimental and Statistical Hypotheses Estimating Variances When There Are More Than Two Conditions in Your Experiment Assumptions for Analysis of Variance Testing Hypotheses About Differences Among Population Means With Analysis of Variance Factors That Affect the Power of the F-Test in Analysis of Variance Relational Effect Size Measures for Analysis of Variance Randomization Tests for Testing for Differential Effects of Three or More Treatments Using ANOVA to Study the Effects of More Than One Factor on Behavior Partitioning Variance for a Two-Factor Analysis of Variance Testing Hypotheses With Two-Factor Analysis of Variance Testing Hypotheses About Differences Among Population Means With Analysis of Variance Dealing With Unequal Sample Sizes in Factorial Designs Summary Conceptual Exercises Chapter 15: Multiple Regression and Beyond Overview of the General Linear Model Approach Regression Simple Versus Multiple Regression Multiple Regression Types of Multiple Regression Interactions in Multiple Regression Continuous x Continuous Interactions Categorical x Continuous Interactions Categorical x Categorical Interactions: ANOVA Versus Regression Summary Conceptual Exercises Epilogue Appendices A. Some Useful Rules of Algebra B. Rules of Summation C. Logarithms D. The Inverse of the Cumulative Normal Distribution E. The Unit Normal Distribution F. The t-Distribution G. The Fisher r to zr Transformation H. Critical Values for F With Alpha = .05 I. The Chi Square Distribution References Index