Question

Robust linear regression using the t model: The folder congress has the votes for the Democratic and Republican candidates in each U.S. congressional district in 1988, along with the parties’ vote proportions in 1986 and an indicator for whether the incumbent was running for reelection in 1988. For your analysis, just use the elections that were contested by both parties in both years. (a) Fit a linear regression (with the usual normal-distribution model for the errors) predicting 1988 Democratic vote share from the other variables and assess model fit. (b) Fit a t-regression model predicting 1988 Democratic vote share from the other variables and assess model fit; to fit this model in R you can use the tlm() function in the hett package. (See the end of Section C.2 for instructions on loading R packages.) (c) Which model do you prefer?

          Robust linear regression using the t model: The folder congress has the votes for the Democratic and Republican candidates in each U.S. congressional district in 1988, along with the parties’ vote proportions in 1986 and an indicator for whether the incumbent was running for reelection in 1988. For your analysis, just use the elections that were contested by both parties in both years.
(a) Fit a linear regression (with the usual normal-distribution model for the errors) predicting 1988 Democratic vote share from the other variables and
assess model fit.
(b) Fit a t-regression model predicting 1988 Democratic vote share from the other variables and assess model fit; to fit this model in R you can use the tlm() function in the hett package. (See the end of Section C.2 for instructions on loading R packages.)
(c) Which model do you prefer?

Added by Arwell Nathán L.

Question

Please give Ace some feedback