Inference about the Regression Parameters Using Median-Ranked Set Sampling

• Published in: Communications in statistics. Theory and methods Vol. 39; no. 14; pp. 2604 - 2616
• Main Authors: Alodat, M. T., Al-Rawwash, M. Y., Nawajah, I. M.
• Format: Journal Article
• Language: English
• Published: Philadelphia Taylor & Francis Group 01.01.2010; Taylor & Francis Ltd
• Subjects: Asymptotic distribution; Berry-Esseen; Data collection; Inference; Median-ranked set sampling; Primary 62D05; Regression; Regression analysis; Sampling; Secondary 62F05; Simple linear regression; Statistics; Studies
• ISSN: 0361-0926; 1532-415X
• DOI: 10.1080/03610920903072416

The ranked set samples and median ranked set samples in particular have been used extensively in the literature due to many reasons. In some situations, the experimenter may not be able to quantify or measure the response variable due to the high cost of data collection, however it may be easier to rank the subject of interest. The purpose of this article is to study the asymptotic distribution of the parameter estimators of the simple linear regression model. We show that these estimators using median ranked set sampling scheme converge in distribution to the normal distribution under weak conditions. Moreover, we derive large sample confidence intervals for the regression parameters as well as a large sample prediction interval for new observation. Also, we study the properties of these estimators for small sample setup and conduct a simulation study to investigate the behavior of the distributions of the proposed estimators.