Home Membership

R Programs for Estimation of Conditional Logit ParametersThis section contains programs in R that compute both model parameters by maximum likelihood estimate (MLE) and Nash equilibrium strategies for unified models of party competition using conditional logit. Versions are available for univariate (one dimensional) and multivariate (multidimensional) spatial models and for proximity, mixed proximitydirectional, and discounting assumptions. All models permit a partyid variable; the multivariate models also permit additional nonpolicy variables. The template programs assume five parties, 100 respondents; the multivariate versions assume three policy variables and two nonpolicy variables (partyid and income). These parameters can be changed by the user as needed. All programs and data sets assume mean placement of parties by respondents unless otherwise indicated.REFERENCE: A Unified Theory of Party Competition, by James Adams, Samuel Merrill, III, and Bernard Grofman (Cambridge, 2005). See especially Chapter 2 for model definitions. DOWNLOADING AND RUNNING R: R (such as R2.14.0) is open source (no charge) and can be downloaded from http://cran.at.rproject.org/. As indicated on the webpage just mentioned, it is preferred that R be downloaded from a mirror (site) near the user. A list of sites is provided from a link on the webpage. An introductory manual is also available. To run any of the R programs below, open R and copy and paste the program into R after saving the corresponding data file in a text file (in a working folder). The R programs can also be edited as needed in a text editor such as Notepad. R SCRIPTS UNIVARIATE MODELS: Univariate Proximity Unified Model Univariate Mixed Prox Direct Unified Model Univariate Discounting Unified Model Univar Disc Unified Model Idiosyncratic Placement Univar Disc Unif Model Idio Placement & Interaction MULTIVARIATE MODELS:
Multivariate Prox Unified Model Multivariate Mixed Prox Dir Unified Model Multivariate Discounting Unif Model SAMPLE DATA: Univariate data with idiosyncratic placement


Last update: Saturday, May 19, 2012 at 8:28:35 PM. 