Trend function hypothesis testing in the presence of serial correlation

Article Abstract:

Test statistics were developed to validate hypothesis concerning parameters of the deterministic trend function of a univariate time series. The test statistics can be highly effective under the existence of general forms of serial correlation in the errors. They can be utilized without resorting to parametric or nonparametric estimation of serial correlation parameters. Testing revealed the appropriateness of asymptotic approximations on sample sizes that are usually employed in economics.

Economics, Research and Development in the Social Sciences and Humanities, Research, Econometrics, Linear models (Statistics), Parameter estimation

---

Constructing instruments for regressions with measurement error when no additional data are available, with an application to patents and R&D

Article Abstract:

An analysis of two staged least squares used in the estimation of measurement errors in variables of linear regression models reveals that estimates of third moments of the data may be used to identify and estimate errors in the absence of other data. The study reveals that moment estimations are dependent in the skewness of the variables and has useful applications in patenting where aggregate data may not be available.

Models, Statistics, Statistics (Data), Estimation theory, Instrumental variables (Statistics), Instrumental variables, Missing observations (Statistics)

---

Similar abstracts:

• Abstracts: Robust rank tests of the unit root hypothesis. Inference concerning the number of factors in a multivariate nonparametric relationship
• Abstracts: Exact inference methods for first-order autoregressive distributed lag models. Semiparametric latent variable model estimation with endogenous or mismeasured regressors

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.