Testing for Functional Form and Heteroscedasticity
This paper considers tests of misspecification in a heteroscedastic transformation model. We derive Lagrange multiplier (LM) statistics for (i) testing functional form and heteroscedasticity jointly, (ii) testing functional form in the presence of heteroscedasticity, and (iii) testing heteroscedasti...
Saved in:
| Main Author: | |
|---|---|
| Format: | text |
| Language: | English |
| Published: |
Institutional Knowledge at Singapore Management University
2004
|
| Subjects: | |
| Online Access: | https://ink.library.smu.edu.sg/soe_research/836 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Singapore Management University |
| Language: | English |
| Summary: | This paper considers tests of misspecification in a heteroscedastic transformation model. We derive Lagrange multiplier (LM) statistics for (i) testing functional form and heteroscedasticity jointly, (ii) testing functional form in the presence of heteroscedasticity, and (iii) testing heteroscedasticity in the presence of data transformation. We present LM statistics based on the expected information matrix. For cases (i) and (ii), this is done assuming the Box-Cox transformation. For case (iii), the test does not depend on whether the functional form is estimated or pre-specified. Small-sample properties of the tests are assessed by Monte Carlo simulation, and comparisons are made with the likelihood ratio test and other versions of LM test. The results show that the expected-information based LM test has the most appropriate finite-sample empirical size |
|---|