Hardest thing--fundamental understanding of meaning of confidence interval and significance test.
Most people have the mechanics under control.   If not, please see me or Amanda or someone!
- - - - - - - - - - - - - - - - - - - - - - - - - - -  - - - -  - - -
#3) Find z* for 92% CI:   Meant to put into practice the picture you memorized for the quiz.    C is 92%.  To use the z table, you have to figure out what's in the tail(s).  1-.92 = .8  Divide by 2, find 4% in each tail.
92%+4% = 96% to the left of +z*.  Use the table backward, closest is .9599, for z* = 1.75.
OR, Use the table backward, looking up .0400,  and finding -z* = -1.75.

I also gave part credit if you interpolated in the t-table, finding the number 2/5 of the way between the values for 90% and 95%.  The correct answer was 1.771, close to the z-value, since the normal curve can be approximated by a straight line in that narrow interval (1.645 to 1.960).
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
#4e)  A test has been done, attempting to show that lower-income children have lower sentence complexity than middle-income children.  The true mean sentence complexity for middle income children is 15.55
So H₀: µ=15.55     64 low-income children give xbar = 7.62, and we assume sigma =8.91 (which is s)
     H_a: µ<15.55      The z value, given that the null hypothesis is true, is -7.1
Is this strong evidence that low income children have lower sentence complexity?....
Yes (assuming the 64 can be considered a SRS of low income children.)
Why?  An outcome that would rarely happen if a claim were true is good evidence that the claim is not true.
  The xbar is more than 7 standard deviations from the mean, if the mean is really 15.55.  That is super-rare--way off our tables for z .  The chances of seeing such an outcome are way less than .0003, the last value in the table.
This outcome is in the direction of our alternative hypothesis, so it is strong evidence that the null is false--alternative is true.
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
#10  "90% CI for the mean size of all U.S. gardens is 561 to 724 sq ft."  Looking for understanding of CI definition.
Doesn't mean 90% of all gardens in U.S. are between 561 and 724 sq ft.
Doesn't mean that 90% of the time the mean size will fall between 561 and 724 (the next CI you make will have different end points.)
Doesn't mean that there is 90% probability that 561--724 captures the true mean. (It either does or it doesn't, once we've calculated it.  We don't know which; but our definition of probability doesn't apply to experiments that have already happened.)

Means that the Gallup Organization claims that the true but unknown mean size (average of all gardens) is between 561 and 724.   They have "90% confidence" in their claim, which means that they use a method which guarantees  that 90% of all such "90% confidence interval" claims that they make are actually true. (But they/we don't know which are true and which are false.)
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -  -
These and the rest are written up (maybe a bit differently) in the solutions.