Skip to content

FRM Part I · FRM Exam Part I · Regression with Multiple Explanatory Variables

Which remedy is most appropriate for severe, but not perfect, multicollinearity between two regressors that both have theoretical justification in the model?

The suitable remedy is to gather more data or merge the correlated regressors into one composite variable, accepting some loss of detail. White's robust errors fix heteroskedasticity rather than collinearity, and dropping all variables or using adjusted R-squared would not resolve it.

  1. AObtain more data or combine the variables into a single composite, accepting some loss of detailCorrect
  2. BDrop both variables and re-estimate using only the intercept
  3. CUse the adjusted R-squared to select the larger coefficient
  4. DApply White's robust standard errors to eliminate the problem

Explanation

Common remedies include collecting more data, combining correlated variables into an index, or dropping one redundant variable. Robust standard errors address heteroskedasticity, not collinearity. Adjusted R-squared does not remove the problem, and dropping both discards all explanatory content.

Did you get it right without looking?

One question tells you little. A timed set on Regression with Multiple Explanatory Variables shows your real accuracy, how long you take and where you lose marks.

Practise Regression with Multiple Explanatory VariablesFree diagnostic mock

More Regression with Multiple Explanatory Variables questions