A Note on Identification in the Multinomial Probit Model

By Michael P. Keane. 1992

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Outsourcing Household Production: Foreign Domestic Workers and Native Labor Supply in Hong Kong (2013, background): “While a normalized multivariate probit is formally identified as long as the model includes at least one variable that varies at the individual level ðHeckman and Sedlacek 1985Þ, in the absence of exclusion restrictions, the model is extremely fragile ðKeane 1992Þ.”
Multinomial probit and multinomial logit: a comparison of choice models for voting research (2004, methodology): “In this case, the MNP model may produce arbitrary parameter estimates within the tolerance of the estimation procedure (Keane, 1992; Alvarez and Nagler, 1998).”
On the Use of Factor-Analytic Multinomial Logit Item Response Models to Account for Individual Differences in Response Style (2010, methodology): “This phenomenon has already been documented for models that are related to FAMLMs including structural equation models (Kenny, 1979) and multinomial probit models (Keane, 1992; Weeks, 1997).”
Using Heteroskedastic Ordered Probit Models to Recover Moments of Continuous Test Score Distributions From Coarsened Data (2017, background): “In addition, ML algorithms can sometimes indicate convergence even when the multinomial or ordinal probit model is formally unidentified, due to approximation errors in estimating the likelihood function (Horowitz, Sparmann, & Daganzo, 1982; Keane, 1992).”
Federal Reserve Bank of Minneapolis Research Department Staff Report 237 Mixture of Normals Probit Models (unknown): “In view of the irregular likelihood surface in the dimensions of the variance matrix in this model (Keane 1992), whether full or scale mixture of models will succeed as generalizations of the multinomial probit model seems to us very much an open question.”
Invited Paper - Learning Models: An Assessment of Progress, Challenges, and New Developments (2011, unknown): “Keane (2010a, b), and Rust (2010). These authors argue that drawing inferences from data always relies on some set of maintained assumptions.”
The Composite Marginal Likelihood (CML) Inference Approach with Applications to Discrete and Mixed Dependent Variable Models (2014, methodology): “Also, in MNP models, identification is tenuous when only individual-specific covariates are used (see Keane, 1992 and Munkin and Trivedi, 2008).”
Quasi-structural estimation of a model of childcare choices and child cognitive ability production (2010, methodology): “In forming the likelihood function, we simulate the MNP choice probabilities using the GHK probability simulator (see Keane (1994)).”
A New Generalized Heterogeneous Data Model (GHDM) to Jointly Model Mixed Types of Dependent Variables (2015, methodology): “An early example of such a multivariate model may be found in Keane (1992), who considered one nominal variable and one continuous variable.”
Employment, Hours of Work and the Optimal Taxation of Low Income Families (2008, methodology): “Given the difficulty in identifying flexible correlation structures from observed outcomes (see Keane, 1992), we allow ǫy to be correlated with ǫw, but otherwise assume that the errors are independent.”
A New Estimation Approach to Integrate Latent Psychological Constructs in Choice Modeling (2014, methodology): “Early examples include the studies by Revelt and Train (1996), Bhat (1997), and Bhat (1998), and there have now been many applications of this approach, using latent multinomial logit and mixed logit formulations.”
Multilevel logistic regression for polytomous data and rankings (2003, background): “Unfortunately, models only including unit specific covariates are prone to empirical underidentification even in this case, but these problems appear to be reduced when unit and alternative specific covariates x a i j are included (e.g., Keane, 1992).”