liv-class


This class is used to store and further analyze the results of the liv function


Fits linear models with endogenous regressors using Internal
Instrumental Variables (IIV) methods. These are statistical techniques to
correct for endogeneity when no strong, valid external instrumental variables
are available. The first version of the package offers two methods, the Latent
Instrumental Variables (Ebbes et al., 2005) and Lewbel's higher moments approach
(Lewbel, 1997). In a second version of the package, two other methods will be
added, joint estimation using copulas (Park and Gupta, 2012) and multilevel GMM
(Kim and Frees, 2007).

Raluca Gui

REndo

Fitting Linear Models with Endogenous Regressors when No
External Instruments are Available

liv-class function

 <dl>
<dt><code>formula</code></dt><dd>returns an object of type 'formula' as used in the call of the function. Example <code>var1 ~ var2</code>.</dd> <dt><code>coefficients</code></dt><dd>model's coefficients</dd> <dt><code>seCoefficients</code></dt><dd>the standard errors of the coefficients</dd> <dt><code>groupMeans</code></dt><dd>the coefficients of the means of the two groups considered to build the latent IV.</dd> <dt><code>seMeans</code></dt><dd>the standard errors of the groups means coefficients.</dd> <dt><code>sigma</code></dt><dd>the coefficients of the variance -covariance matrix.</dd> <dt><code>probG1</code></dt><dd>the coefficient of the probability of group 1.</dd> <dt><code>seProbG1</code></dt><dd>the standard error of the coefcients of probability of group 1.</dd> <dt><code>initValues</code></dt><dd>the initial parameter values.</dd> <dt><code>value</code></dt><dd>the value of the log=likelihood function computed at the optimal parameter values.</dd> <dt><code>convCode</code></dt><dd>the converge code.</dd> <dt><code>hessian</code></dt><dd>the hessian matrix.</dd> </dl>

liv-class: liv S4 Object

Description

Arguments

Slots

Examples