Abstract
We provide a new characterization of the equality of two positive-definite matrices A and B, and we use this to propose several new computationally convenient statistical tests for the equality of two unknown positivedefinite matrices. Our primary focus is on testing the information matrix equality (e.g. White, 1982, 1994). We characterize the asymptotic behavior of our new trace-determinant information matrix test statistics under the null and the alternative and investigate their finite-sample performance for a variety of models: linear regression, exponential duration, probit, and Tobit. The parametric bootstrap suggested by Horowitz (1994) delivers critical values that provide admirable level behavior, even in samples as small as n =50. Our new tests often have better power than the parametric-bootstrap version of the traditional IMT; when they do not, they nevertheless perform respectably.
| Original language | English |
|---|---|
| Pages (from-to) | 491-556 |
| Number of pages | 66 |
| Journal | Advances in Econometrics |
| Volume | 33 |
| DOIs | |
| Publication status | Published - 2014 Jan 1 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Economics and Econometrics
Cite this
}
Testing the equality of two positive-definite matrices with application to informationmatrix testing. / Cho, Jin Seo; White, Halbert.
In: Advances in Econometrics, Vol. 33, 01.01.2014, p. 491-556.Research output: Contribution to journal › Article
TY - JOUR
T1 - Testing the equality of two positive-definite matrices with application to informationmatrix testing
AU - Cho, Jin Seo
AU - White, Halbert
PY - 2014/1/1
Y1 - 2014/1/1
N2 - We provide a new characterization of the equality of two positive-definite matrices A and B, and we use this to propose several new computationally convenient statistical tests for the equality of two unknown positivedefinite matrices. Our primary focus is on testing the information matrix equality (e.g. White, 1982, 1994). We characterize the asymptotic behavior of our new trace-determinant information matrix test statistics under the null and the alternative and investigate their finite-sample performance for a variety of models: linear regression, exponential duration, probit, and Tobit. The parametric bootstrap suggested by Horowitz (1994) delivers critical values that provide admirable level behavior, even in samples as small as n =50. Our new tests often have better power than the parametric-bootstrap version of the traditional IMT; when they do not, they nevertheless perform respectably.
AB - We provide a new characterization of the equality of two positive-definite matrices A and B, and we use this to propose several new computationally convenient statistical tests for the equality of two unknown positivedefinite matrices. Our primary focus is on testing the information matrix equality (e.g. White, 1982, 1994). We characterize the asymptotic behavior of our new trace-determinant information matrix test statistics under the null and the alternative and investigate their finite-sample performance for a variety of models: linear regression, exponential duration, probit, and Tobit. The parametric bootstrap suggested by Horowitz (1994) delivers critical values that provide admirable level behavior, even in samples as small as n =50. Our new tests often have better power than the parametric-bootstrap version of the traditional IMT; when they do not, they nevertheless perform respectably.
UR - http://www.scopus.com/inward/record.url?scp=84912565684&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=84912565684&partnerID=8YFLogxK
U2 - 10.1108/S0731-905320140000033014
DO - 10.1108/S0731-905320140000033014
M3 - Article
AN - SCOPUS:84912565684
VL - 33
SP - 491
EP - 556
JO - Advances in Econometrics
JF - Advances in Econometrics
SN - 0731-9053
ER -