BIOL 373 Lecture Notes - Lecture 17: Simple Linear Regression, Linear Regression, Biostatistics
Document Summary
Considers a predictive relationship between two variable: dependent & Simple linear regression straight line, a linear relationship. Y =a+bx: linear regression for a population: : error (residual departure of yi from the value of y predicted. F ration: variance due to regression / error variance. Regression line is compared to a background level of variation. F ratio: variance due to factor / error variance. Variation can be distributed in different parts (y i y )2 k i=1 ni j=1 (x ij x)2 n i=1. Regression ss + error ss n i=1 n i=1 (^y i y )2 (y i ^y i)2 k: number of levels ni: sample size for ith level. Groups ss + error ss k i=1 k i=1 ni( x i x)2 ni j=1 ni(x ij x i)2. Y : total variation in y o: the regression accounts for variation in y.