Generalized Class of Estimators Using Dual of Auxiliary Variable Under Non-Response

The present study considers the problem of estimating the population mean of study variable by making the dual use of an auxiliary information in the presence of non-response. We have proposed an improved general family of estimators using an auxiliary variable as well as its rank under simple random sampling without replacement (SRSWOR) and two-phase sampling scheme. We have shown the effective performance of proposed class of estimators over usual unbiased, ratio, regression and difference estimators, Singh and Kumar's (Austral N Z Stat 50(4):395-402, 2008) estimator, Riaz and Shabbir (Pak J Stat, 30(1):83-100, 2014), Yaqub et al.'s (Commun Stat Theory Methods 46(24):12130-12151, 2017) estimator and Irfan et al.'s (Sci Iran, 2020. https://doi.org/10.24200/SCI.2020.53592.3318) estimator under SRSWOR and the works: Singh and Kumar's (Stat Oper Res Trans 33(1):71-84, 2009; Stat Pap 51:559-402, 2010), Kumar's (Commun Stat Appl Methods 19(5):663-671, 2012), Riaz and Shabbir's (Pak J Stat, 30(1):83-100, 2014), Yaqub et al.'s (Commun Stat Theory Methods 46(24):12130-12151, 2017) and Irfan et al.'s (Sci Iran, 2020. https://doi.org/10.24200/SCI.2020.53592.3318) estimators under two-phase sampling using an empirical study. A simulation study is also carried out in support of our objectives.