Hi All,

The College board recently released the 2012 AP Statistics Free Response questions today. Possible solutions, 2012 FR questions.

Any corrections, comments, and feedback appreciated.

Note: I am not a representative of the College Board, nor a member of any committees about the AP exam. I have been a grader in past years, but cannot represent what are complete /correct answers will look like at the grading.

This is just one teacher posting what might be good answers. But tear them up if I screwed up.

### Like this:

Like Loading...

*Related*

## About roughlynormal

I have been a math/statistics teacher for 20 years. I currently teach at a college prep school in southern California. I also coach teaching fellows for Math for America - Los Angeles chapter. I love my career, my colleagues, and my friends & family.

The actual questions seem not to be downloadable at this time so I’m working off memories of seeing them a few days ago.

While agree with your answer in question 5 (the flabby town question), I believe there is another legitimate answer. If a reputable study was done, and 35% of people came up as fit, then our best guess of how many people who are fit would be …. well … 35%.

If the town fathers then asked for a statistical one way test to be run showing that number of fit people is less than 35 – and here we are believing that 35% is our best guess, then I’m thinking we’d only have a 5% of rejecting the null hypothesis – which is, after all, what we actually believe at this point.

My apologies if I’ve gotten something minor in the statement of the problem wrong.

I think overall this question is a brilliant check on fragile learning. I was quite disheartened to find that almost non of my students were able to follow the point I was making.

I think I understand the logic you make. But I wonder if this line of reasoning places a lot of trust in not sweating something that should always be checked: was the sample taken from a random sample of the population of interest? If we don’t have this verified, then how can we test any of the numbers claimed ?