Maximum score estimator

<p>In <a href="/facts/Statistics/DNPsKGYU">statistics</a> and <a href="/facts/Econometrics/7U52gzHt">econometrics</a>, the maximum score estimator is a <a href="/facts/Nonparametric/svTDJHTb">nonparametric</a> <a href="/facts/Statistical_estimation/4nbgFDec">estimator</a> for <a href="/facts/Discrete_choice/hSA6DKCX">discrete choice</a> models developed by <a href="/facts/Charles_Manski/NteECQMw">Charles Manski</a> in 1975. Unlike the <a href="/facts/Multinomial_probit/HxwRDglc">multinomial probit</a> and <a href="/facts/Multinomial_logit/aeF5lqDE">multinomial logit</a> estimators, it makes no assumptions about the <a href="/facts/Probability_distribution/EpsKKVRu">distribution</a> of the unobservable part of <a href="/facts/Utility_theory/wlMDG74L">utility</a>. However, its statistical properties (particularly its <a href="/facts/Asymptotic_distribution/dN7jVHqm">asymptotic distribution</a>) are more complicated than the multinomial probit and logit models, making <a href="/facts/Statistical_inference/sJLjubxm">statistical inference</a> difficult. To address these issues, Joel Horowitz proposed a variant, called the smoothed maximum score estimator.
</p>

Maximum score estimator open-in-new

Maximum score estimator