Estimating Variances in Weighted Least-Squares Estimation of Distributional Parameters

Abstract

Many estimation methods have been proposed for the parameters of statistical distribution. The least squares estimation method, based on a regression model or probability plot, is frequently used by practitioners since its implementation procedure is extremely simple in complete and censoring data cases. However, in the procedure, heteroscedasticity is present in the used regression model and, thus, the weighted least squares estimation or alternative methods should be used. This study proposes an alternative method for the estimation of variance, based on a dependent variable generated via simulation, in order to estimate distributional parameters using the weighted least squares method. In the estimation procedure, the variances or weights are expressed as a function of the rank of the data point in the sample. The considered weighted estimation method is evaluated for the shape parameter of the log-logistic and Weibull distributions via a simulation study. It is found that the considered weighted estimation method shows better performance than the maximum likelihood, least-squares, and certain other alternative estimation approaches in terms of mean square error for most of the considered sample sizes. In addition, a real-life example from hydrology is provided to demonstrate the performance of the considered method.

Toward the Prediction of Bed Load Rating Curve Parameter Values: The Influence of Scale, Particle Size, and Entrainment Threshold

“…It should also be acknowledged that r is strictly applicable only to linear relationships with constant variance (homoscedasticity) over the range of the data (Nagelkerke, ). Various goodness‐of‐fit measures for nonlinear models and/or heteroscedastic data have been proposed, but no consensus on the best procedures has emerged (Cameron, ; Kantar, ; Zhang, ). We assessed the linearity and variance structure of our data by visually inspecting plots of measured versus predicted transport rates.…”

Role of human capital in the promotion of technological progress, economic growth and development in Africa: A case study of Kenya

“…The GLS estimates are equivalent to applying ordinary LS to a linearly transformed form of the data. The transformed form of the data is always uncorrelated, with constant variance (Kantar 2015). The transformed variables are used in weighted multiple linear regression (WREG) to estimate WREG.…”

“…As a result, the covariance matrices of the dependent variable of these models are not in the form σ 2 I, but of the form σ 2 T = ∑, where σ 2 is unknown and T is known (Engeman & Keefe 1982;Kantar 2015;White 1969). Thus, the LS estimates of the coefficients may not have minimum variance.…”

“…, and, consequently, the transformed data set is uncorrelated, with constant variance (Kantar 2015).…”

Efficient estimation of Pareto model using modified maximum likelihood estimators

“…First example is taken from Clark [28] and it was also used by Kantar [29], it consists of 21 observations on the data about number of deaths in major earthquakes during 1900-2011 as published by the U.S. Geological Survey.…”