PENGAMBILAN SAMPEL BERDASARKAN PERINGKAT PADA ANALISIS REGRESI LINIER SEDERHANA

Abstract

Ranked Set Sampling and Ranked Set Sampling concomitant are more efficient than Simple Random Sampling. This can be determined by calculating the Relative Precision which is a ratio value from the variance of the mean from each sampling technique. From the research of Ranked Set Sampling, obtained Var(Y ̅_[m]r')<Var(Y ̅'), RP'>1 and RS'>0 so Ranked Set Sampling is more efficient than Simple Random Sampling. For the research of Ranked Set Sampling concomitant, obtained Var(Y ̅_X(m)r')<Var(Y ̅'), 〖RP〗_Y^c'>1, 〖RS〗_Y^c'>0 so Ranked Set Sampling concomitant is more efficient than Simple Random Sampling, and for simple linear regression analysis obtained Var(β ̂_0(RSS) ')<Var(β ̂_0(SRS) '), Var(β ̂_1(RSS) ')<Var(β ̂_1(SRS) '), 〖RP〗_(β ̂_0 )'>1, 〖RP〗_(β ̂_1 )'>1 so simple linear regression model of Ranked Set Sampling is more efficient than simple linear regression model of Simple Random Sampling. Keywords : Ranked Set Sampling, Ranked Set Sampling Concomitant, Relative Precision, Simple Linear Regression Analysis