yigasq

Hey there,

I have been going through the exercises and had hard time grasping the exercise 7.53. I think it's because of percentages and I couldn't work backwards looking at the answers. Can someone help me a bit, for example how do we plug in a random sample of 1%? Or any further?

Thanks.

David

For reference, this is one of the most difficult exercises in the textbook. :)

It's probably easiest to ignore 1.0% it's a percentage, because the percentage is just a unit of reporting in this case (that is, this is not intended as a subtle proportion context, which might be a first instinct). Suppose you had wanted to detect a difference of 1.0. How small must the SE be for this to be achievable with 80% power? The figure in Example 7.37 on page 282 might be helpful (that example is walking through an analogous case, so it is worth reviewing it in greater detail if you're stuck).

Once you get your number, you can put the "%" unit back on the result. You should get ~0.36%. Note that the solutions are slightly off as is the 2.03 numbers (should be 1.97) in part (c) of the exercise question -- see the typos spreadsheet for the book, linked from openintro.org/book/os, for full details.

Best,
David