Determining model Adequacy for a Trivariate Seemingly Unrelated Regression Model
| dc.contributor.author | Okewole, Dorcas Modupe | |
| dc.date.accessioned | 2022-03-01T09:14:49Z | |
| dc.date.available | 2022-03-01T09:14:49Z | |
| dc.date.issued | 2018 | |
| dc.description.abstract | The Seemingly Unrelated Regression (SUR) estimator was used in a Monte Carlo experiment. Data sets were generated for samples sizes: N=10, 20, 50 and 100. Adjusted Wald ( ) and Wald ( ) were used to measure the goodness-of-fit and linked to , in comparison with other goodness-of-fit measures (McElroy, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC)) for model adequacy in variable selection using the backward elimination procedure. The independent variables and the random disturbances were tested to confirm their distributions using Shapiro-Wilk test of normality. Pearson correlation matrix was used to establish the contemporaneous correlation between the error terms in the equations and the pairwise correlations between the independent variables. The five measures were tested with the generated data and the results showed that the convergence of and measures was faster than that of McElroy , and gave similar results as AIC and BIC. and measures remain unchanged when the extreme (irrelevant) variables were eliminated but the sharp change in the measure was obvious when one relevant variable was eliminated from the model. The convergence of McElroy was very slow compared with the other four measures and the change in the measure was not visible when one of the relevant variables was eliminated. The Adjusted Wald ( ) and Wald ( ) are adequate measures for determining Seemingly Unrelated Regression models in the scenarios for this study. | en_US |
| dc.identifier.issn | 1596-9649 | |
| dc.identifier.uri | http://dspace.run.edu.ng:8080/jspui/handle/123456789/1506 | |
| dc.language.iso | en | en_US |
| dc.publisher | International Research and Development Institute, Nigeria | en_US |
| dc.relation.ispartofseries | Volume 15 Number 1; | |
| dc.subject | Seemingly Unrelated Regression | en_US |
| dc.subject | McElroy | en_US |
| dc.subject | Model Adequacy | en_US |
| dc.subject | Goodness of fit | en_US |
| dc.title | Determining model Adequacy for a Trivariate Seemingly Unrelated Regression Model | en_US |
| dc.type | Article | en_US |