Search

In this paper, we proposed new conventional unbiased and calibration estimators for estimating finite population variance for a Mail Survey Design characterized by the presence of non-response in practice. The properties of the proposed estimators are studied theoretically and numerically. Empirical studies were conducted using two each of existing and simulated data to illustrate the performance of proposed estimators over existing ones. The results of both theoretical and numerical comparison depicted the superiority of proposed estimators for estimating population variance for a Mail Survey Design characterized by non-response. [ABSTRACT FROM AUTHOR]

In this paper, we have proposed a regression-type compromised imputation methods free of unknown parameters. The properties (biases and MSEs) of the proposed class of estimators are derived up to first order approximation using Taylor series approach. Also, the conditions for which the proposed estimators are more efficient than other estimators considered in the study were established. Results of numerical illustration using both real and simulated data revealed that the proposed estimators are more efficient and practicable than exiting estimators considered in the study. [ABSTRACT FROM AUTHOR]

In this paper, we have proposed six classes of estimators for different cases of non-response for a heterogeneous population under the stratified random sampling using auxiliary information with unknown population mean. The properties of the proposed estimators have been examined and studied for fixed sample size and the members of the proposed classes are also defined. Also, a numerical study has been included in support of the enhanced efficiency of the proposed estimator. [ABSTRACT FROM AUTHOR]

In this paper, we have proposed two log- product -type estimators and a new estimator for estimation of finite population mean under measurement error by using auxiliary information. The expressions for Bias and mean squared error of proposed estimators are evaluated up to first order of approximation. Based on theoretical results obtained, a numerical study by generating Normal population using R programming language is also included to compare the efficiency of proposed estimators with other relevant estimators. [ABSTRACT FROM AUTHOR]

In this paper, we suggest an exponential type estimator for estimating population mean under simple random sampling without replacement and also utilize this estimator for missing data under varied imputation techniques. Expressions for Bias and MSE’s are acquired in the form of population parameters up to the first order of approximation. The empirical study in support of theoretical results is also included. [ABSTRACT FROM AUTHOR]

Use of auxiliary information is always suggested at the planning and estimation stage to make the estimators perform more eggiciently. Estimation using auxiliary information is common in sampling literature but using distribution of study and auxiliary information at the estimation stage is uncommon and found useful especially when dealing with rare variable. This study utilizes the auxiliary information and Poisson distributed variates for proposing the log-type estimator and another generalized estimator for finite population mean under simple random sampling without replacement. The Mean Square Error expressions of the proposed estimators are obtained.It is revealed from empirical (point estimation and interval estimation) & theoretical study that use of log type estimators along with suitable auxiliary information for Poisson distributed variates excels the performance of estimators in terms of efficiency. [ABSTRACT FROM AUTHOR]

In this paper, we propose an exponential type estimator for finite population variance under simple random sampling without replacement using an auxiliary variable, which is highly correlated with the study variable. Mean square error of the proposed estimator is derived up to first order of approximation. Efficiency of the proposed estimator is compared with other existing estimators both theoretically and numerically. Results indicate that the proposed estimator is more efficient than the existing estimators considered here. [ABSTRACT FROM AUTHOR]

In this paper, we have proposed two log- product -type estimators and a new estimator for estimation of finite population mean under measurement error by using auxiliary information. The expressions for Bias and mean squared error of proposed estimators are evaluated up to first order of approximation. Based on theoretical results obtained, a numerical study by generating Normal population using R programming language is also included to compare the efficiency of proposed estimators with other relevant estimators. [ABSTRACT FROM AUTHOR]

In this paper, it is shown that an estimator of the population mean using known coefficient of variation and a preliminary test of significance for the variance σ²y = σ²0 is found to be better than the usual regression estimator. [ABSTRACT FROM AUTHOR]