<p>As lab partners, Sally and Betty collected data for a significance test. Both calculated the same Z statistic, but Sally found the results were significant at the a=.05 level while Betty found that the results were not. When checking their results, the women found that the only difference in their work was that Sally had used a two sided test, while Betty used a one sided test. Which of the following could have been their test statistic?</p>
<p>A) -1.980
B) -1.690
C) 1.340
D) 1.690
E) 1.780</p>
<p>How do I do this one? I think it's impossible...if Sally used one sided then maybe...</p>
<p>we did this problem in class, so i am SURE that it is correct:</p>
<p>the girls used the (a) value, because it is the value that is furthest away from zero. this is a very conceptual question; the one sided test (because it covers a larger area) only "specializes" on that one side, so the value used has to be much larger in order for the result for the one sided, but not the two sided, to be significant.</p>
<p>it is hard to find the right words to explain this! i hope this helps!</p>
<p>(and i am SURE that it is (a))</p>
<p>to muffin99000 :</p>
<p>you live in Rockland?</p>
<p>normalCdf(.025) vs normCdf(.05) so the test stat falls inbetween these two z values either positive or negative. So you mark Sally as -1.96 and 1.96. Then you have to have a test stat that falls outside this region, but inside either the -1.65 (or more) or 1.65(or less) region. Therefore you know it has to be -1.98 and the other Betty z score thing is 1.65.</p>
<p>Any other questions, post em. I need more practice because my teacher forgot to bring in his old tests.</p>
<p>How about that question where they ask about the smallest confidence interval that would contain 4000 of 10000 something?</p>
<p>If you have it, type it out so we can solve it!</p>