# StratMLE-class: Fitted Strategic Discrete Choice Models In jrnold/r-strat: Statistical Strategic Models

## Description

This class encapsulates results of a maximum likelihood estimation of the parameters of strategic discrete choice model.

## Objects from the Class

Objects can be created by calls of the form new("StratMLE", ...), but most often as a result of a call to strat.

## Slots

All the slots from "mle" are present. In addition, the object contains the following slots.

call:

Object of class "language". The call to strat.

nobs:

Object of class integer. Number of observations.

na.action:

Object of class character.

contrasts:

Object of class lists.

xlevels:

Object of class list. Each element contains the levels used for variables in the corresponding terms.

terms:

Object of class list. Each element is a terms. These are the formulas for the utilities of all (outcome, player) combinations.

model:

Object of class list. Each element is a model.frame.

pinfo:

Object of class list. Each element is numeric. List of the variances on the private utility terms for each (outcome, player) combination.

oPlayers

Object of class character.

oResponses

Object of class character.

alpha

Object of class list.

players

Object of class character. Names of all players in the game. The order of the elements in this vector determines the player's number. For example, object@players[1] is the name of player 1, and so forth.

outcome

Object of class character. Names of all outcomes in the game. This is redundant, as it is the same information as colnames(object@Y)

.

\item

tNodes Object of class character. Each element is the outcome associated with the terminal node. For example, object@tNodes[1] is the name of the response variable of terminal node 1.

\item

paths Object of class list. Each element is a character vector of the names of the actions forming the path to that terminal node. For example, object@paths[1] is the path to terminal node 1.

## Extends

Class "mle", directly.

## Methods

Since "StratMLE" extends "mle" and can use all methods thereof. In addition, the following methods are available.

call

signature(object = "StratMLE"): the estimates of the model parameters.

coef

signature(object = "StratMLE"): Returns the estimates of the model parameters.

coef

signature(object = "logLik"): Returns an object of class logLik, with the log-likelihood of the model.

formula

signature(object = "StratMLE"): Returns a list of the model formula.

terms

signature(object = "StratMLE"): Returns a list of the model terms.

## Note

This class provides a common interface to the mle results all classes extending StratModel and describing specific games.

For a detailed explanation of the estimating procedure, see the references.

For a another implementation of statistical discrete choice models not in R, see STRAT.

## References

Signorino, C. S. (1999) Strategic interaction and the statistical analysis of international conflict ,American Political Science Review 93, 279–297.

Signorino, C. S. (2002) Strategy and selection in international relations, International Interactions 28, 93–115.

Signorino, C. S. (2003) Structure and Uncertainty in Discrete Choice Models, Political Analysis 11, 316–344.

Signorino, C. S. (2003) STRAT: A Program for Analyzing Statistical Strategic Models, Version 1.4, http://www.rochester.edu/college/psc/signorino/.

Signorino, C. S. and Yilmaz, K. (2003) Strategic misspecification in regression models American Journal Of Political Science 47, 551–566.

See also StratMLE() should be used to generate objects of this class, and "StratModel" which this class extends.