Advances and Applications in Statistics
Volume 7, Issue 3, Pages 451 - 462
(December 2007)
|
|
ON ROBUSTNESS AND CHOOSING BETWEEN TWO NONLINEARITIES
I. D. Essi (Nigeria) and J. O. Iyaniwura (Nigeria)
|
Abstract: It is frequent to encounter situations where a choice is to be made between additive-error and multiplicative-error based models. This study focuses on some of the consequences of making a wrong choice, i.e., assuming a multiplicative error based model instead of an additive error based model and vice-versa, in the context of the popular Cobb-Douglas model. In a Monte Carlo study in which the above investigations have been conducted, sample sizes of 20, 40 and 80 are used with each experiment replicated 20 times. It has been found that the extent of robustness of the mis-specified models varies with the sum of the powers of the input variables. Also as the variance of the error term (hence output) increases, the second mis-specification becomes increasingly less robust than the first. These results have a lot of implications and applications, especially in biometrics and econometrics. |
Keywords and phrases: additive error, multiplicative error, Cobb-Douglas model, robustness. |
|
Number of Downloads: 362 | Number of Views: 1084 |
|