Next >

An Introduction to Nonparametric Statistics

Probability Background Probability Distributions for Observations Gaussian Distribution Uniform Distribution Laplace Distribution Cauchy Distribution Logistic Distribution Exponential Distribution Location and Scale Families Sampling Distributions Binomial Distribution ˜2-distribution T-distribution F-distribution Elementary Tasks in Frequentist Inference Hypothesis Testing One-Sided Hypothesis Tests Two-Sided Hypothesis Tests P-values Confidence Intervals P-value Inversion Test Inversion with Pivotal Statistics A Problematic Example Exercises One-Sample Nonparametric Inference Parametric Inference on Means Estimation Using Averages One-Sample Testing for Gaussian Observations The Need for Distribution-Free Tests One-Sample Median Methods Estimates of the Population Median Hypothesis Tests Concerning the Population Median Con dence Intervals for the Median Inference for Other Quantiles Comparing Tests Power, Sample Size, and Effect Size Power Sample and Effect Sizes Effciency Calculations Examples of Power Calculations Distribution Function Estimation Exercises Two-Sample Testing Two-Sample Approximately Gaussian Inference Two-Sample Approximately Gaussian Inference on Expectations Approximately Gaussian Dispersion Inference General Two-Sample Rank Tests Null Distributions of General Rank Statistics Moments of Rank Statistics A First Distribution-Free Test The Mann-Whitney-Wilcoxon Test Exact and Approximate Mann-Whitney Probabilities Moments and Approximate Normality Other Scoring Schemes Using Data as Scores: the Permutation Test Empirical Levels and Powers of Two-Sample Tests Adaptation to the Presence of Tied Observations Mann-Whitney-Wilcoxon Null Hypotheses E ciency and Power of Two-Sample Tests Efficacy of the Gaussian-Theory Test Efficacy of the Mann-Whitney-Wilcoxon Test Summarizing Asymptotic Relative Efficiency Power for Mann-Whitney-Wilcoxon Testing Testing Equality of Dispersion Two-Sample Estimation and Con dence Intervals Inversion of the Mann-Whitney-Wilcoxon Test Tests for Broad Alternatives Exercises Methods for Three or More Groups Gaussian-Theory Methods Contrasts Multiple Comparisons General Rank Tests Moments of General Rank Sums Construction of a Chi-Square-Distributed Statistic The Kruskal-Wallis Test Kruskal-Wallis Approximate Critical Values Other Scores for Multi-Sample Rank Based Tests Multiple Comparisons Ordered Alternatives Powers of Tests Power of Tests for Ordered Alternatives Power of Tests for Unordered Alternatives E ciency Calculations Ordered Alternatives Unordered Alternatives Exercises Group Differences with Blocking Gaussian Theory Approaches Paired Comparisons Multiple Group Comparisons Nonparametric Paired Comparisons Estimating the Population Median Di erence Con dence Intervals Signed-Rank Statistic Alternative Distribution Two-Way Non-Parametric Analysis of Variance Distribution of Rank Sums A Generalization of the Test of Friedman The Balanced Case The Unbalanced Case Multiple Comparisons and Scoring Tests for a Putative Ordering in Two-Way Layouts Exercises Bivariate Methods Parametric Approach Permutation Inference Nonparametric Correlation Rank Correlation Alternative Expectation of the Spearman Correlation Kendall's Bivariate Semi-Parametric Estimation via Correlation Inversion of the Test of Zero Correlation Inversion of the Pearson Correlation Inversion of Kendall's T Inversion of the Spearman Correlation Exercises Multivariate Analysis Standard Parametric Approaches Multivariate Estimation One-Sample Testing Two-Sample Testing Nonparametric Multivariate Estimation Equivariance Properties Nonparametric One-Sample Testing Approaches More General Permutation Solutions Con dence Regions for a Vector Shift Parameter Two-Sample Methods Hypothesis Testing Permutation Testing Permutation Distribution Approximations Exercises Density Estimation Histograms Kernel Density Estimates Exercises Regression Function Estimates Standard Regression Inference Kernel and Local Regression Smoothing Isotonic Regression Splines Quantile Regression Fitting the Quantile Regression Model Resistant Regression Exercises Resampling Techniques The Bootstrap Idea The Bootstrap Sampling Scheme Univariate Bootstrap Techniques The Normal Method Basic Interval The Percentile Method BCa Method Summary So Far, and More Examples Bootstrapping Multivariate Data Sets Regression Models and the Studentized Bootstrap Method Fixed X Bootstrap The Jackknife Examples of Biases of the Proper Order Bias Correction Correcting the Bias in Mean Estimators Correcting the Bias in Quantile Estimators Exercises Appendix A: Analysis Using the SAS System Appendix B: Construction of Heuristic Tables and Figures Using R Bibliography

Next >

Introduction to inferential statistics
Inferential statistics and probability Probability is at the core of inferential statistics and different views about probability (objective/Frequentist vs. subjective) shape different approaches to inferential statistics (Classical/Frequentist vs. Bayesian). Researchers may want to answer questions...
(Statistics for the Behavioural Sciences: An Introduction to Frequentist and Bayesian Approaches)
Statistical Methods: Introduction to Probability Theory
Discrete Variables and Continuous Variables Probability theory deals with both discrete variables and continuous variables. In the case of discrete variables, there are discrete outcomes, which are observed when an event occurs. Examples of events of this kind are the tossing of sets of coins...
(Statistical and Thermal Physics: An Introduction)
An Introduction to Firearms Examination for Researchers in Statistics
Susan VanderPlas, Alicia Carriquiry, Heike Hofmann, James Hamby, and Xiao Hui Tai In the United States, where firearms are readily accessible, the annual number of gun- related crimes is in the hundreds of thousands, and about two thirds of all murders are committed with a gun (Federal...
(Handbook of Forensic Statistics)
A Introduction to Statistical Physics
A.1 Introduction Statistical physics and quantum physics constitute the foundation of the modern physics. They provide methods to study properties of matter. Methods of statistical physics allow us to study macroscopic properties of large systems using microscopic mechanisms and structures proposed...
(Physics of Magnetic Thin Films: Theory and Simulation)
Statistical power analysis: A brief introduction and its application to the one-sample t-test
As explained in Chapter 4, classical inferential statistics is based on obtaining the probability of the certain data given the null hypothesis being true and a significance test procedure, i.e. P(data he null hypothesis and the significance test procedure), with the aim to control and minimise...
(Statistics for the Behavioural Sciences: An Introduction to Frequentist and Bayesian Approaches)
‘Lies, damned lies and statistics’
This chapter is perhaps controversially titled after a much-misquoted phrase, which seems particularly apt when describing the view many academics hold of Higher Education Ranking Systems or HERS for short. Perhaps the earliest citation of the phrase can be attributed to Arthur James Balfour,...
(The Impact of Higher Education Ranking Systems on Universities)
This Book Is an Introduction to Crime Analysis
This book is about crime analysis, which means there must be a theory about crime that will guide the teaching and the approach to the job of analyzing crime and criminal behavior. At this point, we are not ready to put forth that theory. However, we can introduce you to the analysis triangle. This triangle...
(Crime and Intelligence Analysis: An Integrated Real-Time Approach)
Where Do the Statistics for the UCR Program Come From?
In 2019, FBI UCR data were compiled from more than 18,000 law enforcement agencies, representing more than 98% of the population in the United States. Although reporting by law enforcement is not mandated, many states have instituted laws requiring law enforcement within those states to provide UCR data....
(Crime and Intelligence Analysis: An Integrated Real-Time Approach)
Viewing Crime through the Prism of Statistics
While the measures of crime provide us with standardized data on crime, crime rates, and comparative data for states and cities, statistics have their limitations. Perhaps the greatest limitation is that the human factor is missing. Looking at statistics regarding homicides, burglaries, and sexual assault...
(Crime and Intelligence Analysis: An Integrated Real-Time Approach)
Utilizing Statistics as a Form of Accountability: CompStat
CompStat, or comparative statistics, is an accountability process officially adopted by the New York City Police Department in 1994. By and large, CompStat, as used by law enforcement, is an in-house process that holds upper management accountable for crime reduction within their respective areas of...
(Crime and Intelligence Analysis: An Integrated Real-Time Approach)