4. On Prediction: We aim to model food expenditure using the next model FEE: =51+2'?3n0;+61', where FE stands for the food expenditure of the 3'\" household (in $), and nq stands for the weekly income of the 2'\" household (in $100), and 61- is the error term. After regression using 40 observations, we obtain the estimated model as follows. EB = 83.42 + 10.21 - inc, (43.41) (2.09) where the numbers in parentheses are the corresponding standard errors. Calculate the 95% prediction interval for household with income $2000. Note that = 21 and 3 = 89.52, where % stands for the average household income of the dataset, and 3 is the estimated standard error of the error term