What is being done in part i? And why is the statistic less than or equal to o.oo5.. It shoul lie between -0.005 and 0.005.explain!????
0.05* Question demand for minimum "n" to see CI of maximum length 10% of \(\bar X \) i.e. maximum of (0.95\(\bar X \), 1,05\(\bar X \))=(\(\bar X \) - 0.05, \(\bar X \) + 0.05) at 95% level. ................( as \(\bar X \) is proportion. now compare general CI at 95% level with (\(\bar X \) -0.05, \(\bar X \) +0.05). you see!
0.05* actually statistic to CI is lie between -0.05 and 0.05, not just less than 0.05. it was just addition/subtraction to the mean that to be less than 0.05.
Hello all, apologies for bringing back an old thread. But I have a query with regards to this question. IN part (i) solution, they say " Since n is constant it is the p(1-p) which determines the width. The GREATEST value this can take is when p = 0.5." Now mathematically I understand why p(1-p) is the greatest when p = 0.5. But why did they take p =0.5 to begin with? The way I am trying to solve this is via the range method. I end up at the exact equation towards the end of the solution for part (i). But i got stuck at what p-hat should be. If someone could please provide me with an explanation I would be very happy!
You need to show minimum value of 'Dr' for 'Nr/Dr < Constant' be true. Now, you know some range of Numerator. Think, for what extent you can get minimum value of Denominator. If Nr is lowest, Dr for the inequality shows minimum value higher than when Nr is highest. So, we choose Nr highest.