Extra Credit: After-school tutoring programs are a billion-dollar industry. Suppose we have the following regression model estimated by a novice program evaluator (not from Metro State’s MPNA-MPA-MNLM-MS ADC/CODC graduate programs) showing that tutoring programs have a negative effect on student grades:
student grade^ = 3.2 – 1.0 (tutoring program)
(t* = - 6.91)
Explain why the estimated effect of tutoring program (coded 1 if the student is participating in a tutoring program, 0 otherwise) in this regression model is biased. Provide at least two omitted antecedent variables that may be biasing downwards the effect of tutoring programs. Discuss. What do you expect the direction of the effect of tutoring programs on learning outcomes to be after controlling for these antecedent variables? What will happen to the t-statistic (t*) in terms of absolute value and sign after including and controlling for these variables in the regression model? Support your answer.