A Novel Estimator for Finite Population Mean in the Presence of Minimum and Maximum Values

The search for an efficient estimator of the finite population mean has been a critical problem to the sample survey research community. This study is motivated by the fact that the conducted literature review showed that no research has developed such an average ratio estimator of the population mean that would utilize both the population and the sample medians of study variable, as well as the Srivastava (1967) estimator at a time. In this paper we proposed the power ratio cum median-based ratio estimator of the finite population mean, which is a function of two ratio estimators in the form of an average. The estimator assumes the population to be homogeneous and skewed. The properties (i.e. the Bias and the Mean Squared Error – MSE) of the proposed estimator were derived alongside its asymptotically optimum MSE. We demonstrated the efficiency of the proposed estimator jointly with its efficiency conditions by comparing it to selected estimators described in the literature. Empirically, a real-life dataset from the literature and a simulation study from two skewed distributions (Gamma and Weibull) were used to examine the efficiency gain. The empirical analysis and simulation study demonstrated that the efficiency gain is significant. Hence, the practical application of the proposed estimator is recommended, especially in socio-economic surveys.

In this paper we study the joint treatment of not missing at random response mechanism and informative sampling for survey data. This is the most general situation in surveys and other combinations of sampling informativeness and response mechanisms can be considered as special cases. The proposed method combines two methodologies used in the analysis of sample surveys for the treatment of informative sampling and the nonignorable nonresponse mechanism. One incorporates the dependence of the first order inclusion probabilities on the study variable, while the other incorporates the dependence of the probability of nonresponse on unobserved or missing observations. The main purpose here is the estimation of finite population mean and superpopulation parameters when the sampling design is informative and nonresponse mechanism is nonignorable. Under four scenarios of sampling design and nonresponse mechanism, we obtained the method of moment estimators of finite population mean, with their biases and mean square errors. Furthermore, a four-step estimation method is introduced for the estimation of superpopulation parameters under informative sampling and nonignorable nonresponse mechanism. New relationships between moments of response, nonresponse, sample, sample-complement and population distributions were derived. Most estimators for finite population mean known from sampling surveys can be derived as a special case of the results derived in this paper.

An improved family of estimators of finite population mean based on the auxiliary attribute

Accurate estimation of the finite population mean is a fundamental challenge in survey sampling, especially when dealing with large or complex populations. Traditional methods like simple random sampling may not always provide reliable or efficient estimates in such cases. Motivated by this, the current study explores complex sampling techniques to improve the precision and accuracy of mean estimators. Specifically, we employ two-stage and three-stage cluster sampling methods to develop unbiased estimators for the finite population mean. Building upon these, the next phase of the study formulates unbiased mean estimators using stratified two- and three-stage cluster sampling. To further enhance the precision of these estimators, a ranked-set sampling strategy is applied to the secondary and tertiary sampling stages. Additionally, unbiased variance estimators corresponding to the proposed mean estimators are derived. Real-world datasets are utilized to demonstrate the application of these complex survey sampling methodologies, with results showing that the mean estimates derived using ranked set sampling are more accurate than those obtained via simple random sampling.

<abstract> This article deals with the estimation of the finite population mean under probability proportional to size (PPS) sampling using information on the auxiliary variable along with the rank of the auxiliary variable. We propose a ratio, product and regression type estimators by incorporating the maximum and minimum values of the study variable and the auxiliary variable. The mathematical expressions of the proposed estimators are derived up to first order of approximation. Efficiency comparisons are made on the basis of real data sets. </abstract>

ABSTRACTWhen a sufficient correlation between the study variable and the auxiliary variable exists, the ranks of the auxiliary variable are also correlated with the study variable, and thus, these ranks can be used as an effective tool in increasing the precision of an estimator. In this paper, we propose a new improved estimator of the finite population mean that incorporates the supplementary information in forms of: (i) the auxiliary variable and (ii) ranks of the auxiliary variable. Mathematical expressions for the bias and the mean-squared error of the proposed estimator are derived under the first order of approximation. The theoretical and empirical studies reveal that the proposed estimator always performs better than the usual mean, ratio, product, exponential-ratio and -product, classical regression estimators, and Rao (1991), Singh et al. (2009), Shabbir and Gupta (2010), Grover and Kaur (2011, 2014) estimators.

This paper is intended as an investigation of constructing almost unbiased estimators of finite population mean by suitably combining a set of transformed estimators. A generalization of Tracy, Singh and Singh (1999) estimator is suggested, and, to the first degree of approximation, each member of the proposed class is as efficient as the usual regression estimator. Further, it is proved that Reddy (1973) estimator is also a particular case of the proposed class. To the second degree of approximation, a new almost unbiased estimator is established. Moreover, an empirical study is carried out in order to understand better the performance of the new estimator compared to the usual unbiased B, ratio BR and Tracy et al. (1999) estimators.

In literature, there are many estimators of finite population mean, some of which are superior to the others. In practical situations, all the information on sample units may not be available due to non-response in sample surveys. Thus, our objective in this study is to get more precise estimators of the finite population mean of the study variable under two-phase sampling in case of missing data. Three new logarithmic ratio cum logarithmic product type imputation methods and corresponding point estimators have been introduced and observed to be better under two-phase sampling, which adds contribution to the field of imputation techniques. The bias and mean square errors of the proposed estimators are calculated in terms of population parameters. The performance of the proposed estimators is compared theoretically and empirically as well with existing traditional estimators.

In this paper we consider the problem of estimation of population mean using information on two auxiliary variables in systematic sampling. We have extended Singh (1967) estimator for estimation of population mean in systematic sampling. We have derived the expressions for the bias and mean squared error of the suggested estimator up to the first degree of approximation. We have compared the suggested estimator with existing estimators and obtained the conditions under which the suggested estimator is more efficient. An empirical study has been carried out to demonstrate the performance of the suggested estimator.

In this study, we propose generalized exponential type ratio-cum-ratio estimators of finite population mean using ranked Set Sampling (RSS) and stratified ranked set sampling (SRSS) schemes. The biases and Mean Square Errors (MSEs) of the proposed estimators are obtained up to first degree of approximation. Comparisons among the proposed and some existing estimators are made both theoretically and through simulation studies. Also, the estimators are compared in terms of relative bias (RB), relative mean square error (RRMSE) and percentage relative efficiency (PRE). It turned out that when the variable of interest and the auxiliary variables jointly follow a trivariate normal distribution, the proposed class of estimators of the population mean dominates all other competitor estimators.

In sampling theory, it is a popular trend to use auxiliary information to obtain more efficient estimators for the population parameters to increase the precision of the estimator. Estimators obtained using auxiliary information are supposed to be more efficient than the estimators obtained without using auxiliary information. The ratio, regression, product and difference methods take advantage of the auxiliary information at the estimation stage. Therefore, this study considered a generalized class of log-type estimators of finite population mean based on correlation coefficient as the proposed estimator for estimating the population mean of the study variable. It has been shown that the generalized class of log-type estimators has lesser mean square errors (MSEs) under the optimum values of the characterizing scalar as compared to some of the commonly used related estimators available in the literature. Further, an extension of the proposed generalized class of log-type estimators using multiple auxiliary variables such as coefficient of variation, coefficient of kurtosis , and correlation coefficient. have also base initiated in this dissertation. The expressions for the properties of the proposed family of estimators, that is; Bias and Mean Square Error (MSE), were derived to the first degree of approximation. We also obtained the optimum Mean Square Error (MSEopt.), and theoretical comparisons were made with the related existing estimators in literature. Following theoretical comparisons, it was demonstrated that the proposed family of estimators was more efficient than various related existing estimators compared with, under the obtained conditions.

ABSTRACTThis study focuses on the estimation of population mean of a sensitive variable in stratified random sampling based on randomized response technique (RRT) when the observations are contaminated by measurement errors (ME). A generalized estimator of population mean is proposed by using additively scrambled responses for the sensitive variable. The expressions for the bias and mean square error (MSE) of the proposed estimator are derived. The performance of the proposed estimator is evaluated both theoretically and empirically. Results are also applied to a real data set.

The purpose of this study is to present a generalized class of estimators using the three-stage optional randomized response technique (RRT) in the presence of non-response and measurement errors on a sensitive study variable. The proposed estimator makes use of dual auxiliary information. The expression for the bias and mean square error of the proposed estimator are derived using Taylor series expansion. The proposed estimator's applicability is proven using real data sets. A numerical study is used to compare efficiency of the proposed estimator with adapted estimators of the finite population mean. The suggested estimator performs better than adapted ordinary, ratio, and exponential ratio-type estimators in the presence of both non-response and measurement errors. The efficiency of the proposed estimator of population mean declines as the inverse sampling rate, non-response rate, and sensitivity level of the survey question increase.

The purpose of this study is to present a generalized class of estimators using the three-stage Optional Randomized Response Technique (ORRT) in the presence of non-response and measurement errors on a sensitive study variable. The proposed estimator makes use of dual auxiliary information. The expression for the bias and mean square error of the proposed estimator are derived using Taylor series expansion. The proposed estimator’s applicability is proven using real data sets. A numerical study is used to compare the efficiency of the proposed estimator with adapted estimators of the finite population mean. The suggested estimator performs better than adapted ordinary, ratio, and exponential ratio-type estimators in the presence of both non-response and measurement errors. The efficiency of the proposed estimator of population mean declines as the inverse sampling rate, non-response rate, and sensitivity level of the survey question increase.

In this study, two-auxiliary exponential type estimators of finite population mean in Two-phase are proposed. The proposed estimators are extension of (1) estimator finite population mean in SRS to Two-phase sampling. The study investigated efficiency of the proposed estimators by utilizing the ratio of bias to standard error (RBSE) as a proxy to examine confidence limits for estimates. The expressions for the bias and Mean Square Error (MSE) of the estimators were derived. A comprehensive simulation study was carried out to show the efficacy of the estimators as compared to conventional estimators using Coefficient of Variation as a performance measure. Furthermore, a small sample from real data set was utilized to validate the performance of proposed estimators under two varying correlation coefficients amongst variables in the parameter space. The results of both the simulation study and real life studies have shown that the proposed estimators were not only asymptotic, more efficient but produces estimates that are more precise than most of the existing estimators considered in this study.