The Dow Jones Global Indexes describe over 3000 companies worldwide that represent more than 80% of the equity capital in the world’s stock markets. Data file XR17068 shows the quarterly indexes for the world as a whole, and indexes for the United States, the Americas, Europe, and the Asia/Pacific region over 23 consecutive quarters. Given that the dependent variable will be y = world index, construct a correlation matrix for the four possible predictor variables. Does multicollinearity seem to be present in the data? If so, would this present a problem if we were using the predictor variables to estimate the world index? If there is multicollinearity, would this present a problem if we were interpreting the partial regression coefficients in a linear model applied to the data?

  • CreatedSeptember 08, 2015
  • Files Included
Post your question