1. Multicollinearity. Suppose that you postulate a model explaining final exam score in terms of class attendance. Thus, the dependent variable is final exam score, and the key explanatory variable is number of classes attended. To control for student abilities and efforts outside the classroom, you include among the explanatory variables various measures of the student's previous year performance. Someone says, "You cannot hope to learn anything from this exercise because the measures of performance that you included are likely to be highly collinear." What should your response be?

Fig: 1