Diagnostic testing a time series regression with zero-one dummies

Philip Hans Franses; Mikhail Muradov

doi:10.1016/j.econlet.2026.113172

Diagnostic testing a time series regression with zero-one dummies

Philip Hans Franses^*
, Mikhail Muradov

^*Corresponding author for this work

Econometrics

Econometric Institute

Research output: Contribution to journal › Article › Academic › peer-review

1 Downloads (Pure)

Abstract

We consider tests for normality, residual autocorrelation, and autoregressive conditional heteroskedasticity for a time series regression, which includes various zero-one dummy variables for individual observations. The sizes of these tests are distorted as the residuals associated with the dummies are equal to zero. We propose a simple solution which amounts to dismissing the observations with these zero-valued residuals, that is, treat these observations as missing. With simulations we show that the resulting tests have the proper size.

Original language	English
Article number	113172
Journal	Economics Letters
Volume	268
DOIs	https://doi.org/10.1016/j.econlet.2026.113172
Publication status	Published - Sept 2026

Bibliographical note

Publisher Copyright:
© 2026 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license. http://creativecommons.org/licenses/by/4.0/

JEL
C22; C52

Access to Document

10.1016/j.econlet.2026.113172Licence: CC BY

Diagnostic testing a time series regression with zero-one dummiesFinal published version, 509 KBLicence: CC BY

Cite this

@article{6e9915164f9a4ceb8246ec6969198891,

title = "Diagnostic testing a time series regression with zero-one dummies",

abstract = "We consider tests for normality, residual autocorrelation, and autoregressive conditional heteroskedasticity for a time series regression, which includes various zero-one dummy variables for individual observations. The sizes of these tests are distorted as the residuals associated with the dummies are equal to zero. We propose a simple solution which amounts to dismissing the observations with these zero-valued residuals, that is, treat these observations as missing. With simulations we show that the resulting tests have the proper size.",

author = "Franses, \{Philip Hans\} and Mikhail Muradov",

note = "Publisher Copyright: {\textcopyright} 2026 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license. http://creativecommons.org/licenses/by/4.0/ JEL C22; C52",

year = "2026",

month = sep,

doi = "10.1016/j.econlet.2026.113172",

language = "English",

volume = "268",

journal = "Economics Letters",

issn = "0165-1765",

publisher = "Elsevier",

}

TY - JOUR

T1 - Diagnostic testing a time series regression with zero-one dummies

AU - Franses, Philip Hans

AU - Muradov, Mikhail

PY - 2026/9

Y1 - 2026/9

N2 - We consider tests for normality, residual autocorrelation, and autoregressive conditional heteroskedasticity for a time series regression, which includes various zero-one dummy variables for individual observations. The sizes of these tests are distorted as the residuals associated with the dummies are equal to zero. We propose a simple solution which amounts to dismissing the observations with these zero-valued residuals, that is, treat these observations as missing. With simulations we show that the resulting tests have the proper size.

AB - We consider tests for normality, residual autocorrelation, and autoregressive conditional heteroskedasticity for a time series regression, which includes various zero-one dummy variables for individual observations. The sizes of these tests are distorted as the residuals associated with the dummies are equal to zero. We propose a simple solution which amounts to dismissing the observations with these zero-valued residuals, that is, treat these observations as missing. With simulations we show that the resulting tests have the proper size.

UR - https://www.scopus.com/pages/publications/105045768837

U2 - 10.1016/j.econlet.2026.113172

DO - 10.1016/j.econlet.2026.113172

M3 - Article

AN - SCOPUS:105045768837

SN - 0165-1765

VL - 268

JO - Economics Letters

JF - Economics Letters

M1 - 113172

ER -

Diagnostic testing a time series regression with zero-one dummies

Abstract

Bibliographical note

Access to Document

Other files and links

Fingerprint

Cite this