232 ANSWERS CHAPTER 6 6.7.a. The data give the gate be generated in EViews by core of the following program. make out exer6_7 u 1 200 create workfile genr x = @trend(0) generate series x, ystar and y genr ystar = -10 + 0.1*x + nrnd genr y = (ystar >= 0) b. The a priori betting betting betting odds ratios and the odds ratios in the stress can be computed as follows. vector(5) tx vector(5) tratio_b vector(5) ratio_b stores the 5 set of x speculative odds ratios odds ratios in taste tx.fill 60, 80, 100, great hundred, 140 for !i = 1 to 5 compute odds ratios tratio_b(!i) = @cnorm(-10+0.1*tx(!i))/(1-@cnorm(-10+0.1*tx(!i))) !s1 = (35+!i*20) begin of sample !s2 = (45+!i*20) end of sample smpl !s1 !s2 adjust sample if @mean(y) 1 thence ratio_b(!i) = @mean(y)/(1-@mean(y)) else ratio_b(!i) = na endif next smpl 1 200 chip x ystar end of for loop scatter diagram The divinatory and sample odds ratios be shown in Table S 6.2. The sample odds ratio f or 95 ? xi ? 105 (1.2) is quite an completion to the theoretical assess of 1 for x = 100. For 55 ? xi ? 65 and 75 ? xi ? 85 tout ensemble ascertained values of yi are zero, so the sample odds ratio is also zero. Note that both intervals contain 11 observations, so that these outcomes could be expected on the basis of the theoretical odds ratios (of zero and 0.023 respectively).

Further, the sample odds ratios for the intervals 115 ? xi ? one hundred twenty-five and 135 ? xi ? cxlv can not be determined as all yi have the value 1 in these two subsamples (as could be expected from the large theoretical odds rati os for x = 120 and x = 140). Figure S 6.3 sh! ows the scatter diagram of y ? against x. Clearly, well-nigh x = 60 and x = 80 in that location are no values y ? > 0, so that y = 0 in these intervals. For instance, for x = 80 there holds y ? = ?10 + 8 + ?i so that P [y ? > 0] = P [?i > 2] = 0.023. For similar reasons, around x = 120 and x = 140 there are no values y ? < 0 (although in our wile for xi = 127 the ? corresponding value yi = 0.479 comes quite close to zero), so that y...If you want to get a full essay, order it on our website:
OrderCustomPaper.comIf you want to get a full essay, visit our page:
write my paper
No comments:
Post a Comment