Specification and testing of hierarchical ordered response models with anchoring vignettes

William Greene; Mark  Harris; Rachel Knott; Nigel Rice

Specification and testing of hierarchical ordered response models with anchoring vignettes

William Greene, Mark Harris , Rachel Knott, Nigel Rice

Research output: Working paper › Discussion paper

Abstract

Anchoring vignettes have been proposed as a way to correct for differential item functioning when individuals self-assess their health, or other aspects of their circumstances on an ordered categorical scale. The model relies on two key underlying assumptions of response consistency and vignette equivalence. Adopting a modified specification of the boundary equations
in the compound hierarchical ordered probit model this paper develops joint and separate tests of these assumptions based on a score approach. Monte Carlo simulations show that the tests have good size and power properties in finite samples. We provide an application of the test to data from the Survey of Health, Aging and Retirement in Europe (SHARE), using self-reported data on pain. The tests are easy to implement, only
requiring estimation of the restricted model under the null hypothesis.

Original language	English
Publisher	NYU/STERN
Number of pages	43
Volume	EC-19-05
Publication status	Published - Aug 2019

Access to Document

http://w4.stern.nyu.edu/economics/docs/workingpapers/2018/VignettetestingGreeneHarrisKnottRice.pdf

Cite this

@techreport{783d8945b8b44c2593fbcda1fd1bd1e4,

title = "Specification and testing of hierarchical ordered response models with anchoring vignettes",

abstract = "Anchoring vignettes have been proposed as a way to correct for differential item functioning when individuals self-assess their health, or other aspects of their circumstances on an ordered categorical scale. The model relies on two key underlying assumptions of response consistency and vignette equivalence. Adopting a modified specification of the boundary equationsin the compound hierarchical ordered probit model this paper develops joint and separate tests of these assumptions based on a score approach. Monte Carlo simulations show that the tests have good size and power properties in finite samples. We provide an application of the test to data from the Survey of Health, Aging and Retirement in Europe (SHARE), using self-reported data on pain. The tests are easy to implement, onlyrequiring estimation of the restricted model under the null hypothesis.",

author = "William Greene and Mark Harris and Rachel Knott and Nigel Rice",

year = "2019",

month = aug,

language = "English",

volume = "EC-19-05",

publisher = "NYU/STERN",

type = "WorkingPaper",

institution = "NYU/STERN",

}

TY - UNPB

T1 - Specification and testing of hierarchical ordered response models with anchoring vignettes

AU - Greene, William

AU - Harris , Mark

AU - Knott, Rachel

AU - Rice, Nigel

PY - 2019/8

Y1 - 2019/8

N2 - Anchoring vignettes have been proposed as a way to correct for differential item functioning when individuals self-assess their health, or other aspects of their circumstances on an ordered categorical scale. The model relies on two key underlying assumptions of response consistency and vignette equivalence. Adopting a modified specification of the boundary equationsin the compound hierarchical ordered probit model this paper develops joint and separate tests of these assumptions based on a score approach. Monte Carlo simulations show that the tests have good size and power properties in finite samples. We provide an application of the test to data from the Survey of Health, Aging and Retirement in Europe (SHARE), using self-reported data on pain. The tests are easy to implement, onlyrequiring estimation of the restricted model under the null hypothesis.

AB - Anchoring vignettes have been proposed as a way to correct for differential item functioning when individuals self-assess their health, or other aspects of their circumstances on an ordered categorical scale. The model relies on two key underlying assumptions of response consistency and vignette equivalence. Adopting a modified specification of the boundary equationsin the compound hierarchical ordered probit model this paper develops joint and separate tests of these assumptions based on a score approach. Monte Carlo simulations show that the tests have good size and power properties in finite samples. We provide an application of the test to data from the Survey of Health, Aging and Retirement in Europe (SHARE), using self-reported data on pain. The tests are easy to implement, onlyrequiring estimation of the restricted model under the null hypothesis.

M3 - Discussion paper

VL - EC-19-05

BT - Specification and testing of hierarchical ordered response models with anchoring vignettes

PB - NYU/STERN

ER -