Question

For our estimates of the population parameters to be consistent in a multivariable regression model, we need which of the following assumptions to be true?

a)There are no large outliers in X1, X2, …, Xk, or Y.

b)There is no perfect multicollinearity.

c)The variance of the error, u, given every value of every X, is constant.

d)There is no imperfect multicollinearity.

e)The sample is randomly selected from the population of interest.

f)The errors, u, are normally distributed. g) E[u | X1,X2,…,Xk]=0

Answer 1

The following assumptions that NEED to be true for the multivariate regression model are as follows:

1. There are no large outliers in X1, X2, …, Xk, or Y.

There should be no outliers present in the data.

2.There is no perfect multicollinearity.

There should be no perfect multicollinearity in the model. Having imperfect multicollinearity is also a violation, but still is accepted up to some extent.

3. The variance of the error, u, given every value of every X, is constant.

The variance of the error term should be constant.

4. The errors, u, are normally distributed.

Error terms should follow a normal distribution.

5. E[u | X1,X2,…,Xk]=0

Mean of error terms should be zero.