STA302H5 Lecture 19: Lecture 19 - 20131113_112520.pdf
Document Summary
Error t value pr(>|t|) (intercept) 12. 964384 0. 3650854 35. 510547 2. 167687e-57 c1 -2. 785812 0. 9101021 -3. 060989 2. 853779e-03 c2 -2. 279768 0. 9390140 -2. 427832 1. 703327e-02. > # first replicate test of h0: beta1=0. > betahat = justcountry; betahat (intercept) c1 c2. > v = vcov(justcountry) # mse * (x"x)-inverse. > t1 = t(a1) %*% betahat / sqrt(t(a1) %*% v %*% a1) > # now test h0: beta1 = beta2. Conclusion: american cars are getting fewer kilometers per litre on average than. > jc2 = aov(lper100k~cntry); summary(jc2) # aov is a wrapper for lm. Df sum sq mean sq f value pr(>f) > # the factor cntry has dummy vars built in. Error t value pr(>|t|) (intercept) 10. 1786 0. 8337 12. 209 < 2e-16 *** > # you can select the dummy variable coding scheme. > contr. treatment(3,base=2) # category 2 is the reference category. Error t value pr(>|t|) (intercept) 12. 9644 0. 3651 35. 511 < 2e-16 *** Residual standard error: 3. 119 on 97 degrees of freedom.
