Statistical Methods for Field and Laboratory Studies in Behavioral Ecology

Statistical Methods for Field and Laboratory Studies in Behavioral Ecology focuses on how statistical methods may be used to make sense of behavioral ecology and other data. It presents fundamental concepts in statistical inference and intermediate topics such as multiple least squares regression and ANOVA. The objective is to teach students to recognize situations where various statistical methods should be used, understand the strengths and limitations of the methods, and to show how they are implemented in R code. Examples are based on research described in the literature of behavioral ecology, with data sets and analysis code provided. Features: This intermediate to advanced statistical methods text was written with the behavioral ecologist in mind Computer programs are provided, written in the R language. Datasets are also provided, mostly based, at least to some degree, on real studies. Methods and ideas discussed include multiple regression and ANOVA, logistic and Poisson regression, machine learning and model identification, time-to-event modeling, time series and stochastic modeling, game-theoretic modeling, multivariate methods, study design/sample size, and what to do when things go wrong. It is assumed that the reader has already had exposure to statistics through a first introductory course at least, and also has sufficient knowledge of R. However, some introductory material is included to aid the less initiated reader. Scott Pardo, Ph.D ., is an accredited professional statistician (PStat®) by the American Statistical Association. Michael Pardo is a Ph.D . is a candidate in behavioral ecology at Cornell University, specializing in animal communication and social behavior. Table of Contents Acknowledgements Preface Statistical Foundations Binary Results – Single Samples and 2 x 2 Tables Continuous Variables – One and Two Samples The Linear Model – Continuous Regressor Variables The Linear Model – Discrete Regressor Variables The Linear Model – Random Effects and Mixed Models Polytomous Discrete Variables – R x C Contingency Tables The Generalized Linear Model – Logistic and Poisson Regression Multivariate Continuous Variables – Dimension Reduction, Clustering, Discrimination Bayesian and Frequentist Philosophies Decision and Game Theory Modern Prediction Methods and Machine Learning Models Time-to-Event Modeling Time Series Analysis and Stochastic Processes Study Design and Sample Size Considerations When Things Go Wrong… APPENDIX. Solving