THE PURPOSE OF THIS PAPER is threefold: (i) to provide a unified iterative generalized least squares approach for computing FIML estimates for both linear and nonlinear simultaneous equations models of medium size (containing about 50 parameters in the nonlinear case and 75 in the linear case); (ii) to suggest a useful procedure to determine the length of the step to be made at each iteration of the algorithm; (iii) to report on numerical results which illustrate the robustness and the relative efficiency of the proposed method.

The threshold regression assumes that value of dependent variable remains fixed until the concerted action of independent variables and error term induces it to overcome its reaction threshold. The existence of reaction thresholds in purchasing of private cars has been discussed and analyzed. Among main explanatory variables used in micro-economic studies, most significant appear to be household composition, age and education of head, number of earners, number and age of cars, value of stock of cars previously held, liquid assets, net worth, debts and place of residence. The independent variables finally retained in present study are permanent income, education of head household and number of children. To summarize, income appears to be most important variable affecting car purchases by households. There are no doubt other social and demographic characteristics of households which bear on this phenomenon, but nature of these variables as well as mechanisms through which they operate have not been sufficiently investigated and are not yet sufficiently well known to permit us to detect their effect very clearly, in statistical investigations.

The invariance properties of some well known asymptotic tests are studied. Three types of invariance are considered: invariance to the representation of the null hypothesis, invariance to one-to-one transformations of the parameter space (reparameterizations), and invariance to one-to-one transformations of the model variables such as changes in measurement units. Tests that are not invariant include the Wald test and generalized versions of it, a widely used variant of the Lagrange multiplier test, Neyman's C(a) test, and a generalized version of the latter. For all these tests, we show that simply changing measurement units can lead to vastly different answers even when equivalent null hypotheses are tested. This problem is illustrated by considering regression models with Box-Cox transformations on the variables. We observe, in particular, that various consistent estimators of the information matrix lead to test procedures with different invariance properties. General sufficient conditions are then established, under which the generalized C(a) test becomes invariant to reformulations of the null hypothesis and/or to one-to-one transformations of the parameter space as well as to transformations of the variables. In many practical cases where Wald-type tests lack invariance, we find that special formulations of the generalized C(a) test are invariant and hardly more costly to compute than Wald tests. This computational simplicity stands in contrast with other invariant tests such as the likelihood ratio test. We conclude that noninvariant asymptotic tests should be avoided or used with great care. Further, in many situations, the suggested implementation of the generalized C(a) test often yields an attractive substitute to the Wald test (which is not invariant) and to other invariant tests (which are more costly to perform).

To facilitate maximum likelihood estimation for Box-Cox models, several authors have suggested dividing the dependent variable by its sample geometric mean. This paper points out previously unmentioned drawbacks of this 'recalling.' First, the 'resealed' model is not actually equivalent to the untransformed one, so that the procedure involves more than a unit change. Second, there is no clear interpretation of the parameters after such resealing. The authors suggest an interpretation but find that the usual formulas for standard errors and confidence intervals are not asymptotically valid. Only tests for zero coefficients are valid. Thirdly, the authors discuss the appropriate way of measuring elasticities in such models.