Binary and Ordered Response Models in Randomized Experiments: Applications of the Resampling-Based Maximum Likelihood Method

Takahiro Ito
Additional contact information
Takahiro Ito: Graduate School of International Cooperation Studies, Kobe University

No 42, GSICS Working Paper Series from Graduate School of International Cooperation Studies, Kobe University

Abstract: This paper formulates a novel distribution-free maximum likelihood estimator for binary and ordered response models and demonstrates its finite sample performance in a Monte Carlo simulation. The simulation examines an ordered response model, focusing on estimating the effect of an exogenous regressor (e.g., randomly assigned treatment status) on the choice probability for an ordered outcome. Estimations are implemented based on a binary specification, which converts the outcome to dichotomous values {0, 1}, or an ordinal specification, which uses the outcome as is. The simulation results show that the proposed estimator outperforms conventional parametric/semiparametric estimators in most cases for both specifications. The results also show that the superiority of the proposed estimator holds even in the presence of conditionally heteroscedastic variance. In addition, the estimates based on the ordinal specification are always superior to those based on the binary specification in all simulation designs, implying that converting ordered responses to dichotomous responses and estimating based on the binary specification may not be the optimal approach.

Keywords: semiparametric estimation; distribution-free maximum likelihood; binary choice model; ordered response model; Likert-type data; heteroscedastic variance (search for similar items in EconPapers)
Pages: 22 pages
Date: 2024-01
New Economics Papers: this item is included in nep-dcm, nep-ecm and nep-inv
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://da.lib.kobe-u.ac.jp/da/kernel/0100486169/ (application/pdf)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:kcs:wpaper:42

Access Statistics for this paper

More papers in GSICS Working Paper Series from Graduate School of International Cooperation Studies, Kobe University Contact information at EDIRC.
Bibliographic data for series maintained by GSICS Library (gsics-library@office.kobe-u.ac.jp).