Academic Integrity: tutoring, explanations, and feedback — we don’t complete graded work or submit on a student’s behalf.

ECON2900Y nt and dependent variables n A researcher is analysing the relationshi

ID: 1105283 • Letter: E

Question

ECON2900Y nt and dependent variables n A researcher is analysing the relationship between the independe sngaession model. She has estimated her first model from a samplie of 204 observations 5% level of significance: s,s = 4 RA2 -0.65 bles Also, she has developed a correlation matrix for her three independent varia Xs -0.1 0.1 1.0 Correlation Matrix Xt -0.1 -0.1 -0.1 1.0 0.1 Xy Based on the correlation results, she has then applied the forward selection search procedure based not only on significance but also on goodness of fit. She believes good element in building models. Using a 5% level of significance she has selected her final model from a sample of 203 observations: ness of fit is the key Y =10+10x, + 10x2 S01 = 1 RA2 = 0.90 a) Was a search procedure necessary? Why? b) Forecast the value of Y when X1 = 10, X2-10 and X: 10? c) Forecast the value of Y when X 1 and X2 1? d) Can you explain 90% of the behavour of Y, using X, X2 and Xs altogether? e) Can you explain 65% of the behaviour of Y, using Xi and X2 altogether? 20 points

Explanation / Answer

a) search process is necessary as it allow us to get the best model and help us decide the relevant variables neede. Thus saving from omitted variable bias as well as the problem of inclusion of unnecessary variable.

b) Y = 10-2*(10)- 4*(10) -8*(10) = -110

c) Y = 10+10*(1)+10*(1) =30

d) Yes, because R2 is 90% which tell 90% of the variation in Y is explained by variation in X's