see images

Suppose you estimate the linear regression model y = Bo + 81@ + e and you find the estimates (bo, b1 ). Let * be some hypothetical value of the independent variable, and let y* be the associated value of the dependent variable. The quantity y* = bo + bja* is then an estimate of Oy O E(y* ) Oy O E(y)Suppose you estimate the linear regression model 3,- = n + 319: + c and you nd the estimates {50:51} Let as\" be some hypothetical value of the independent variable. and let 1," be the associated value of the dependent yariable. To construct a condence interval for Eff }, we use a t distribution With _____ degrees of freedom. OnE Suppose you estimate the linear regression model y = Bo + 812 + e and you find the estimates (bo, b1 ). Let x* be some hypothetical value of the independent variable, and let y* be the associated value of the dependent variable. Suppose you calculate confidence intervals for y* and E(y* ), using the same confidence level. Which of the following is true? (Hint: see Figure 14.11) O The confidence interval for y* is narrower than the interval for E(y*). O The confidence intervals are the same. O The confidence interval for y* is wider than the interval for E(y* ). O The confidence interval for y* could be narrower or wider than the interval for E(y* ).An influential observation is an observation that O is close to the sample means a and y O is far from the sample means a and y O has a high leverage O has a strong influence on the regression results