On the Use of Probit-Based Models for Ranking Data Analysis

Gopindra S. Nair, Chandra R. Bhat, Ram Pendyala, Becky P.Y. Loo, William H.K. Lam

Research output: Contribution to journalArticle

Abstract

In consumer surveys, more information per response regarding preferences of alternatives may be obtained if individuals are asked to rank alternatives instead of being asked to select only the most-preferred alternative. However, the latter method continues to be the common method of preference elicitation. This is because of the belief that ranking of alternatives is cognitively burdensome. In addition, the limited research on modeling ranking data has been based on the rank ordered logit (ROL) model. In this paper, we show that a rank ordered probit (ROP) model can better utilize ranking data information, and that the prevalent view of ranking data as not being reliable (because of the attenuation of model coefficients with rank depth) may be traced to the use of a misspecified ROL model rather than to any cognitive burden considerations.

Original languageEnglish (US)
JournalTransportation Research Record
DOIs
StatePublished - Jan 1 2019

Fingerprint

Data structures

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Cite this

On the Use of Probit-Based Models for Ranking Data Analysis. / Nair, Gopindra S.; Bhat, Chandra R.; Pendyala, Ram; Loo, Becky P.Y.; Lam, William H.K.

In: Transportation Research Record, 01.01.2019.

Research output: Contribution to journalArticle

Nair, Gopindra S. ; Bhat, Chandra R. ; Pendyala, Ram ; Loo, Becky P.Y. ; Lam, William H.K. / On the Use of Probit-Based Models for Ranking Data Analysis. In: Transportation Research Record. 2019.
@article{e7cf434a2f714ea88397bb61b1e50d25,
title = "On the Use of Probit-Based Models for Ranking Data Analysis",
abstract = "In consumer surveys, more information per response regarding preferences of alternatives may be obtained if individuals are asked to rank alternatives instead of being asked to select only the most-preferred alternative. However, the latter method continues to be the common method of preference elicitation. This is because of the belief that ranking of alternatives is cognitively burdensome. In addition, the limited research on modeling ranking data has been based on the rank ordered logit (ROL) model. In this paper, we show that a rank ordered probit (ROP) model can better utilize ranking data information, and that the prevalent view of ranking data as not being reliable (because of the attenuation of model coefficients with rank depth) may be traced to the use of a misspecified ROL model rather than to any cognitive burden considerations.",
author = "Nair, {Gopindra S.} and Bhat, {Chandra R.} and Ram Pendyala and Loo, {Becky P.Y.} and Lam, {William H.K.}",
year = "2019",
month = "1",
day = "1",
doi = "10.1177/0361198119838987",
language = "English (US)",
journal = "Transportation Research Record",
issn = "0361-1981",
publisher = "US National Research Council",

}

TY - JOUR

T1 - On the Use of Probit-Based Models for Ranking Data Analysis

AU - Nair, Gopindra S.

AU - Bhat, Chandra R.

AU - Pendyala, Ram

AU - Loo, Becky P.Y.

AU - Lam, William H.K.

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In consumer surveys, more information per response regarding preferences of alternatives may be obtained if individuals are asked to rank alternatives instead of being asked to select only the most-preferred alternative. However, the latter method continues to be the common method of preference elicitation. This is because of the belief that ranking of alternatives is cognitively burdensome. In addition, the limited research on modeling ranking data has been based on the rank ordered logit (ROL) model. In this paper, we show that a rank ordered probit (ROP) model can better utilize ranking data information, and that the prevalent view of ranking data as not being reliable (because of the attenuation of model coefficients with rank depth) may be traced to the use of a misspecified ROL model rather than to any cognitive burden considerations.

AB - In consumer surveys, more information per response regarding preferences of alternatives may be obtained if individuals are asked to rank alternatives instead of being asked to select only the most-preferred alternative. However, the latter method continues to be the common method of preference elicitation. This is because of the belief that ranking of alternatives is cognitively burdensome. In addition, the limited research on modeling ranking data has been based on the rank ordered logit (ROL) model. In this paper, we show that a rank ordered probit (ROP) model can better utilize ranking data information, and that the prevalent view of ranking data as not being reliable (because of the attenuation of model coefficients with rank depth) may be traced to the use of a misspecified ROL model rather than to any cognitive burden considerations.

UR - http://www.scopus.com/inward/record.url?scp=85064550470&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85064550470&partnerID=8YFLogxK

U2 - 10.1177/0361198119838987

DO - 10.1177/0361198119838987

M3 - Article

JO - Transportation Research Record

JF - Transportation Research Record

SN - 0361-1981

ER -