The document contains two theoretical exercises on regression analysis:
1. The first exercise asks to compute the expected value, variance, and correlation for the random variable Y, which is defined as a linear combination of a bivariate normal random variable X and a standard normal error term.
2. The second exercise involves a log-linear regression model and asks to test the significance of variables, compute standard errors using the delta method, and test whether slope coefficients are equal using provided estimates and standard errors. It also asks to interpret the effect on Y of a one unit increase in X based on the model.
The document contains two theoretical exercises on regression analysis:
1. The first exercise asks to compute the expected value, variance, and correlation for the random variable Y, which is defined as a linear combination of a bivariate normal random variable X and a standard normal error term.
2. The second exercise involves a log-linear regression model and asks to test the significance of variables, compute standard errors using the delta method, and test whether slope coefficients are equal using provided estimates and standard errors. It also asks to interpret the effect on Y of a one unit increase in X based on the model.
The document contains two theoretical exercises on regression analysis:
1. The first exercise asks to compute the expected value, variance, and correlation for the random variable Y, which is defined as a linear combination of a bivariate normal random variable X and a standard normal error term.
2. The second exercise involves a log-linear regression model and asks to test the significance of variables, compute standard errors using the delta method, and test whether slope coefficients are equal using provided estimates and standard errors. It also asks to interpret the effect on Y of a one unit increase in X based on the model.
1. Let X be a bivariate distribution with mean µ and covariance matrix
Σ, where à ! à ! 1 3 1 µ= and Σ = . 0 1 1 Define Y = bt X + ε, where ε follows a standard normal distribution and is independent of X. Finally, let b = (−1, 2)t .
(a) Compute E(Y )
(b) Compute Var(Y ) (c) Compute the correlation between Y and ε.
2. Consider the regression model
log(yi ) = α + βxi + γzi + εi ,
for 1 ≤ i ≤ n and with εi the error terms. Suppose that β̂ = 1.5
and SE(β̂) = 0.75. Suppose that γ̂ = 4 and SE(γ̂) = 0.75. Moreover, Cov(β̂, γ̂) = 0.2.
(a) Is the variable xi significant?
(b) Is β̂ significantly different from 5? (c) Using the delta-method, compute SE(β̂ 2 ). (d) Test whether H0 : β = γ holds. (e) Compute SE(β̂γ̂) (f) Complete the next sentence: “If x increases with 1 unit, then we expect that y increases with ...........”