Answer Happy

You have the following data from a given response variable: 8.04 11.73 y = -1.84 9.46 7.57 and for one predictor given by the design matrix ſi 4.48 1 6.23 X = 1 0.01 1 5.46 1 2.69 You know that the variances of the errors are not constant and they can be scaled using the vector of weights WT = [1 2 0.5 2 0.5]. Use weighted least squares to calculate the following estimators: the residual variance the estimated parameters B and Var(@) the 95% confidence interval for B1 Does the data provide evidence for a relation between the predictor and the response, i.e. for H1 : B170? R code to construct datasets: y <- c(8.04, 11.73, -1.84, 9.46, 7.57) x <- c(4.48, 6.23, 0.01, 5.46, 2.69) X <- cbind(1, x) W <- c(1, 2, 0.5, 2, 0.5)