Written by the creators of Stata's likelihood maximization features, Maximum Likelihood Estimation with Stata, Third Edition continues the pioneering work of the previous editions. Emphasizing practical implications for applied work, the first chapter provides an overview of maximum likelihood estimation theory and numerical optimization methods. With step-by-step instructions, the next several chapters detail the use of Stata to maximize user-written likelihood functions. Various examples include logit, probit, linear, Weibull, and random-effects linear regression as well as the Cox proportional hazards model. The final chapters describe how to add a new estimation command to Stata. Assuming a familiarity with Stata, this reference is ideal for researchers who need to maximize their own likelihood functions. New ml commands and their functions: constraint: fits a model with linear constraints on the coefficient by defining your constraints; accepts a constraint matrix ml model: picks up survey characteristics; accepts the subpop option for analyzing survey data optimization algorithms: Berndt-Hall-Hall-Hausman (BHHH), Davidon-Fletcher-Powell (DFP), Broyden-Fletcher-Goldfarb-Shanno (BFGS) ml: switches between optimization algorithms; computes variance estimates using the outer product of gradients (OPG)

Table of Contents:
Preface Versions of Stata Notation and Typography THEORY AND PRACTICE The likelihood-maximization problem Likelihood theory The maximization problem Monitoring convergence OVERVIEW OF ml The jargon of ml Equations in ml Likelihood-evaluator methods Tools for the ml programmer Common ml options Maximizing your own likelihood functions METHOD lf The linear-form restrictions Examples The importance of generating temporary variables as doubles Problems you can safely ignore Nonlinear specifications The advantages of lf in terms of execution speed The advantages of lf in terms of accuracy METHODS d0, d1, AND d2 Comparing these methods Outline of method d0, d1, and d2 evaluators Summary of methods d0, d1, and d2 Linear-form examples Panel-data likelihoods DEBUGGING LIKELIHOOD EVALUATORS ml check Using methods d1debug and d2debug ml trace SETTING INITIAL VALUES ml search ml plot ml init INTERACTIVE MAXIMIZATION The iteration log Pressing the Break key Maximizing difficult likelihood functions FINAL RESULTS Graphing convergence Redisplaying output WRITING DO-FILES TO MAXIMIZE LIKELIHOODS The structure of a do-file Putting the do-file into production WRITING ADO-FILES TO MAXIMIZE LIKELIHOODS Writing estimation commands The standard estimation-command outline Outline for estimation commands using ml Using ml in noninteractive mode Advice WRITING ADO-FILES FOR SURVEY DATA ANALYSIS Program properties Writing your own predict command OTHER EXAMPLES The logit model The probit model The normal model: linear regression The Weibull model The Cox proportional hazards model The random-effects regression model The seemingly unrelated regression model APPENDIX A: Syntax of ml APPENDIX B: Likelihood evaluator checklists APPENDIX C: Listing of estimation commands References Author Index Subject Index