Journal of Statistical Research of Iran
Volume:11 Issue: 1, 2015

Nowadays, there is an increasing use of sampling methods in network and spatial populations. Although the most common link-tracing designs such as adaptive cluster sampling and snowball sampling have advantages over conventional sampling designs such as simple random sampling and cluster sampling, these designs still present many drawbacks. Adaptive web sampling is a new link-tracing design that overcomes some of the drawbacks. In this paper, after introducing adaptive web sampling design, an application of this method to estimate total production of olive in Iran using 1382 census data of agriculture is presented. Then by conducting a simulation study the performance of adaptive web sampling in estimating total olive production will be compared with that of general inverse adaptive sampling.

Based on the increasing property of ratio of average reversed hazard rates of two non-negative random variables a new stochastic order for the sake of comparison of two lifetime distributions is proposed. This stochastic order admits some distinguishing properties. In order to illustrate the obtained results, a semi-parametric model which is called reverse hazard power and a mixture model of proportional reversed hazards are taken into account. Some examples are given to explain some facts.

In this paper, at first we derive a family of maximum Tsallis entropy distributions under optional side conditions on the mean income and the Gini index. Furthermore, corresponding with these distributions a family of Lorenz curves compatible with the optional side conditions is generated. Meanwhile, we show that our results reduce to Shannon entropy as $beta$ tends to one. Finally, by using actual data, we compare the maximum Tsallis entropy Lorenz curve with some parametric Lorenz curves.

The ratio-type estimators have been introduced for estimating the mean and total population, but in recent years based on the ratio methods several estimators for population variance have been proposed. In this paper two families of estimators have been suggested and their approximation mean square error (MSE) have been developed. In addition, the efficiency of these variance estimators are compared. We also show that their performance is better than traditional ratio estimator. Finally, a computational approach to find the optimum values of parameters is introduced and by using rice fields data of Amol county of Iran, an empirical study is carried out to show that this approach is the best. In this manner, since the values of estimators MSE are approximated, to obtain the approximately unbiased estimator with minimum variance, average square error ASE is used.

In this paper, we study the performance of estimators of parametersof two-parameter exponential distribution based on upper records.The generalized likelihood ratio (GLR) test was used to generate preliminary test estimator (PTE) for both parameters. We have compared the proposed estimator with maximum likelihood (ML) and unbiased estimators (UE) under mean-squared error (MSE) and Pitman measure of closeness (PMC). Analytical as well as graphical methods are used to show the range of parameter in which PTE performs better than ML and UE. Results demonstrate that in the case of that prior information is not too far from its real value, the PTE is superior in compare with ML and UE based on both MSE and PMC criteria. The results of the paper will be useful in estimation with record data in life testing experiments.

In this paper we introduce a four-parameter generalized Weibull distribution. This new distribution has a more general form of failure rate function. It is more general for modeling than six ageing classes of life distributions with appropriate choices of parameter values, so it can display decreasing, increasing, bathtub shaped, unimodal, increasing-decreasing increasing and decreasing-increasing-decreasing failure rates. The new distribution has also a bimodal density function. The moments are obtained and the method of maximum likelihood is used to estimate the model parameters. Also, the observed information matrix is obtained. Two applications are presented to illustrate the advantage of the proposed distribution.