Authors

• Nader Nematollahi ¹
• Reza Shahi ²

¹ Department of Statistics, Allameh Tabataba’i University, Tehran, Iran.

² Department of Mathematics, Tabriz Branch, Islamic Azad University, Iran.

Abstract

Ranked Set Sampling (RSS) is a statistical method for data collection that leads to more efficient estimators than competitors based on Simple Random Sampling (SRS). We consider testing the correlation coefficient of bivariate normal distribution based on Bivariate RSS
(BVRSS). Under one-sided and two-sided alternatives, we show that the new tests based on BVRSS are more powerful than the usual uniformly most powerful tests based on bivariate SRS. Furthermore, the proposed tests for repeated and unrepeated samples are compared.

Keywords

• Bivariate normal distribution
• bivariate ranked set sampling
• Correlation Coefficient
• Hypothesis testing
• ranked set sampling