Each week, we publish layman’s abstracts of new articles from our prestigious portfolio of journals in statistics. The aim is to highlight the latest research to a broader audience in an accessible format.The article featured today is from the Canadian Journal of Statistics with the full article now available to read here.
Ozturk, O. and Kravchuk, O. (2022), Combining ranking information from different sources in ranked-set samples. Can J Statistics. https://doi.org/10.1002/cjs.11656
As a life scientist using a ranked set sampling design in the field, you may often be concerned with obtaining the best estimation of the population mean or total. In this design, you look at the ranks (or relative positions) of potential sampling units in a small set to make a decision about which units to take to the lab for expensive physical measurements. Each record in your final data set will have two values: a numerical value of the physical measurement and the rank you assigned to that unit in its selection set in the field. To increase the information content of your sample, in addition to the original ranks, you can consider assigning K-1 more ranks for each field sample. These multiple ranks can be derived in several ways. For example, you can take inexpensive and non-destructive measurements on each unit in the field, expecting them to be somewhat correlated to the expensive measurement, and obtain ranks based on this covariate; or you could ask critical colleagues accompanying you to follow your original protocol and assign their own ranks to each of the measured units. As a result, each data point in your data set consists of K ranks and the numerical value from the physical measurement. This paper establishes how these K ranks can be efficiently incorporated in the analysis of the expensive measurements to produce more accurate estimates of the mean. The performance of the new estimator is demonstrated on the assessment of early crop establishment of faba bean, a contribution to the emerging trend of digital sampling in agricultural field research.