#6

I'm confused as to how to calculate the degrees of freedom. I came up with 28 diseases in our regression of logdol loghospdays logmort logyrslost logdisbil. Thus, I calculated the degrees of freedom as (28 - 5 - 1) = 22 based on the results of the regression. What did you all get for the degrees of freedom? Based on 22 as the degrees of freedom, I obtained the following results for #6:

predict yhat,xb
predict h, leverage
predict std_yhat, stdp
predict std_f, stdf
generate cil_yhat = yhat - invttail(28-5-1,.025)*std_yhat
generate ciu_yhat = yhat + invttail(28-5-1,.025)*std_yhat
generate cil_f = yhat - invttail(28-5-1,.025)*std_f
generate ciu_f = yhat + invttail(28-5-1,.025)*std_f
generate cil_sbpf = exp(cil_f)
generate ciu_sbpf = exp(ciu_f)
predict t, rstudent
display invttail(22,.025)
2.0738731
lowess t yhat, bwidth(0.2) msymbol(Oh) mcolor(gs10) clwidth(thick)yline(-2.07 0 2.07)
generate out = t > 1.96 | t < -1.96
tabulate out