PROBIT WITH SPATIAL AUTOCORRELATION

Daniel P. McMillen

Research output: Contribution to journal › Article

Abstract

ABSTRACT. Commonly‐employed spatial autocorrelation models imply heteroskedastic errors, but heteroskedasticity causes probit to be inconsistent. This paper proposes and illustrates the use of two categories of estimators for probit models with spatial autocorrelation. One category is based on the EM algorithm, and requires repeated application of a maximum‐likelihood estimator. The other category, which can be applied to models derived using the spatial expansion method, only requires weighted least squares.

Original languageEnglish (US)
Pages (from-to)335-348
Number of pages14
JournalJournal of Regional Science
Volume32
Issue number3
DOIs
StatePublished - Aug 1992

Fingerprint

autocorrelation
cause
method

ASJC Scopus subject areas

  • Development
  • Environmental Science (miscellaneous)

Cite this

PROBIT WITH SPATIAL AUTOCORRELATION. / McMillen, Daniel P.

In: Journal of Regional Science, Vol. 32, No. 3, 08.1992, p. 335-348.

Research output: Contribution to journal › Article

McMillen, DP 1992, 'PROBIT WITH SPATIAL AUTOCORRELATION', Journal of Regional Science, vol. 32, no. 3, pp. 335-348. https://doi.org/10.1111/j.1467-9787.1992.tb00190.x
McMillen DP. PROBIT WITH SPATIAL AUTOCORRELATION. Journal of Regional Science. 1992 Aug;32(3):335-348. https://doi.org/10.1111/j.1467-9787.1992.tb00190.x
McMillen, Daniel P. / PROBIT WITH SPATIAL AUTOCORRELATION. In: Journal of Regional Science. 1992 ; Vol. 32, No. 3. pp. 335-348.
@article{eba62bb1a8b24909a33133015197d276,
title = "PROBIT WITH SPATIAL AUTOCORRELATION",
abstract = "ABSTRACT. Commonly‐employed spatial autocorrelation models imply heteroskedastic errors, but heteroskedasticity causes probit to be inconsistent. This paper proposes and illustrates the use of two categories of estimators for probit models with spatial autocorrelation. One category is based on the EM algorithm, and requires repeated application of a maximum‐likelihood estimator. The other category, which can be applied to models derived using the spatial expansion method, only requires weighted least squares.",
author = "McMillen, {Daniel P.}",
year = "1992",
month = "8",
doi = "10.1111/j.1467-9787.1992.tb00190.x",
language = "English (US)",
volume = "32",
pages = "335--348",
journal = "Journal of Regional Science",
issn = "0022-4146",
publisher = "Wiley-Blackwell",
number = "3",

}

TY - JOUR

T1 - PROBIT WITH SPATIAL AUTOCORRELATION

AU - McMillen, Daniel P.

PY - 1992/8

Y1 - 1992/8

N2 - ABSTRACT. Commonly‐employed spatial autocorrelation models imply heteroskedastic errors, but heteroskedasticity causes probit to be inconsistent. This paper proposes and illustrates the use of two categories of estimators for probit models with spatial autocorrelation. One category is based on the EM algorithm, and requires repeated application of a maximum‐likelihood estimator. The other category, which can be applied to models derived using the spatial expansion method, only requires weighted least squares.

AB - ABSTRACT. Commonly‐employed spatial autocorrelation models imply heteroskedastic errors, but heteroskedasticity causes probit to be inconsistent. This paper proposes and illustrates the use of two categories of estimators for probit models with spatial autocorrelation. One category is based on the EM algorithm, and requires repeated application of a maximum‐likelihood estimator. The other category, which can be applied to models derived using the spatial expansion method, only requires weighted least squares.

UR - http://www.scopus.com/inward/record.url?scp=84985287431&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84985287431&partnerID=8YFLogxK

U2 - 10.1111/j.1467-9787.1992.tb00190.x

DO - 10.1111/j.1467-9787.1992.tb00190.x

M3 - Article

AN - SCOPUS:84985287431

VL - 32

SP - 335

EP - 348

JO - Journal of Regional Science

JF - Journal of Regional Science

SN - 0022-4146

IS - 3

ER -

Access to Document