What is multicollinearity? Explain with an example.

Distinguish near-perfect multicollinearity from perfect multicollinearity, indicating possible consequences in each case. Provide examples.

Define variance inflation factor (VIF)? Calculate its minimum and maximum possible values, interpreting the meaning of VIF in each case.

Suppose we obtain the following OLS estimates of the price of a desktop PC labelled as PCPRICE:

Dependent variable: PCPRICE

VariableCoefficientStd. Errort-StatisticProb. C2373.701189.960012.495790.0000DD-0.0485770.079362-0.6120900.5529VEL-0.3590001.768456-0.2030020.8428RAM8.1294289.1602840.8874650.3938 R-squared0.758659 Mean dependent var2058.990Adjusted R-squared0.692839 S.D. dependent var682.8771S.E. of regression378.4651 Akaike info criterion14.93330Sum squared resid1575594. Schwarz criterion15.12212Log likelihood-107.9998 Hannan-Quinn criter.14.93129F-statistic11.52622 Durbin-Watson stat0.989149Prob(F-statistic)0.001012

The correlation matrix between the regression variables is as follows:

PCPRICEDDVELRAMPCPRICE 1.000000-0.860457-0.860734-0.843369DD-0.860457 1.000000 0.998015 0.995173VEL-0.860734 0.998015 1.000000 0.993130RAM-0.843369 0.995173 0.993130 1.000000

Comment on the estimates of DD, VEL and RAM and also R². Use the correlation matrix above to examine the severity of one potential violation of the Gauss-Markov assumptions in the above Model. Name and define this violation. How can you remedy this?