A linear regression looked at the relationship between the dollar prices P of 45 used 2001 Audi A4 2.8L sedans with manual transmission and the number of miles M the cars have been driven. A least squares regression yields

Suppose that the true relationship is P = α − β₁M + β₂C + ε, where C is the condition of the car and α, β₁, and β₂ are all positive. If M and C are negatively correlated, does the omission of C from the estimated equation bias the estimate of β₁ upward or downward?

Transcribed Image Text:

P = 16,958 -0.0677M, R2 = 0.70 (2,553) (0.0233)