The output below is from the randomized experiment in problem set #4, in which small classes (sck) and regular classes with aides (rak) were randomly assigned to kindergarten students, after which test scores (tscorek) were recorded, in 80 schools. Observation are kindergarten students. summ tscorek sck rak schidk . reg tscorek sck rak, robust Variable Obs Mean Std. Dev. Min Max Linear regression Number of obs 5,748 F(2, 5745) 20.47 tscorek 5,748 922.3904 73.8765 635 1253 Prob > F 0.0000 sck 5,748 . 3014962 . 4589476 1 R-squared 0.0074 rak 5,748 . 3505567 -4771858 Root MSE 73.616 schidkn 5, 748 39.83664 22.95755 80 . reg tscorek sck, robust Robust tscorek Coef. Std. Err. t P>/t| [95% Conf. Interval] Linear regression Number of obs 5,748 F(1, 5746) 40.92 sck 14.10878 2. 458904 5.74 0.000 9.288401 18.92916 Prob > F 0.0000 rak . 5552705 2.28443 0.24 0.808 -3.923073 5.033614 R-squared 0.0074 cons 917.942 1.635777 561.17 0.000 914.7353 921.1487 Root MSE 73.61 . test (sck-0) (rak-0) Robust tscorek Coef. Std. Err. t P>It| [95% Conf. Interval] ( 1) sck - 0 ( 2) rak = 0 5ck 13. 83011 2.16196 6.40 0.000 9.59185 18.06836 F( 2, 5745) = 20.47 Cons 918.2207 1.142023 804.03 0.000 915.9819 920.4595 Prob > F - 0.00301B. In this sample of 5, 748 students, is b21 positive, negative, or zero? Explain . 1C. In a corresponding population of students (possibly infinite), in which sck and rak are randomly assigned, would the omitted variable bias (B2 21) be positive, negative or zero? Explain. 1D. Could the estimated (LR) slopes (14.1 and 0.56) be consistent estimates of linear causal effects, even if b2 were to be nonzero? YES / NO. Explain