Regression Analysis _ Linear

Ordinary least square model (ols) :

• Could be non linear in variables

• Must be linear in the parameters

• Need not be linear in the variables

• All of these

What does the following equation mean : E(ut) =0

• Errors are linearly independents

• The regressor is unrelated to the errors

• Errors have 0 mean

• Erros have variance 0

What is the variable y in the following eq : yt =a + bxt +ut

• Regressor

• Control variable

• predictor variable

• predicted variable

What is the formula for Homoskedasticity

• E(u/x) = E(u)

• E(y/x) = a + bX

• E(u/x) = sigma^2

• None of these

Breush-Pagan is used to test for :

• Omitted Variable

• Homoscedasticity

• Heteroskedasticity

• Proxy variable

From the results obtained, the null hypothesis :

• is rejected at 5% and 1% significance level

• Not rejected at 5% and 1% significance level

• only rejected at 1% significance level

• Not rejected only at 1% significance level

Multicollinearity is used to determine

• The presence of linear relationship among dependent variables

• The presence of linear relationship among endogenous variables

• The presence of non linear relationship among explanatory variables

• The presence of linear relationship among explanatory variables

If for a set of explanatory variables X1 and X2, the coefficient of correlation is equal to 1 , this means that :

• No collinearity

• Low level of collinearity

• Very high collinearity

• Perfect collinearity

A model is obtained as follow by fitting the data : y = 3+ 2X.

The rate of change in the average value of y when the independent value change by 1 unit is :

• 3/2

• 5

• 2

• 3/2

In the model y = 2+3x, the residual corresponding to x=3 and Y =7 is :

• 11

• -4

• 0

• Can not be determined