Subregional mean estimation of agricultural surface area and production under ranked set sampling framework

The aim of this work is to use the data of the Italian Fifth Agricultural Census and of the Farm Structure Survey driven in year 2003, to demonstrate the possibility of the Ranked Set Sampling (RSS) to produce a more precise mean estimation for agricultural surface area and production than the more commonly used sampling design. We are motivated by two reasons, the first being that, although the theoretical benefits of RSS in estimation and statistical inference have been extensively demonstrated and its convenience in agricultural science has been largely emphasized, the methodology has not yet been widely adopted by practitioners. The second one is that the improved precision may allow at territorial level a more detailed estimation than the regional level obtained by the current agricultural survey.