Right now I don’t even know what chi squared is. I’m watching videos now and ill let you know if i need any help thanks!
What type of test do u guys think they are going to ask on the FRQ tomorrow, i am pretty sure paired t-test was last year?
@Gatortristan I’ll give you a general description of chi-squared. For chi-square goodness of fit, the expected values are evenly distributed among the number of observations unless they state the specific percents or ratios. For example, if the problem only state 4 observations, your expected value is 1/4 for each of the 4 observations. The observed value will be given to you in the problem. So, subtract the given observed value to the expected value, square it, and divide it by the expected value. Do this 4 times for the 4 different observations and add them up. This is your chi-squared value. Then you compare it with the table by using the p value of 0.05 and degrees of freedom (your number of observations-1. In this case, it is 4-1=3). Find the critical value in the table and if the one you calculated is bigger than the one in the table, you reject the null hypothesis.
@chienzhouzz
I say a two sample t test, because it been a while since they gave one.
I know Barron’s generally always has irrationally hard practice tests, but I’m getting worried right now. I just got a 2540 on one of the MC and it’s a bit of a confidence crusher, especially since even after I reviewed the solutions, I still don’t know what many of the questions were asking. If any of you use Barron’s stats, how do you guys feel?
@picuberoot I never use barron’s for that reason. They make their tests unjustly hard. Through a few years of AP tests, I found that Princeton Review is the best and usually most accurate. Don’t sweat it, the AP tests are always a lot easier than Barron’s. I’m sure you’re bound for a 4 or a 5.
@picuberoot which edition are you using? I feel like the Barron’s practice test is relatively easy. Anyway, how many questions.
Anyway, how many questions can you miss on MC and still get a 5, assuming that you did decent on the FRQs.
@AnniiT 7th edition. The questions just don’t make sense to me. Also, use appass.com to get a decent gauge on what you need.
I just took the 2012 released exam and scored a 37/39 so I’m feeling a little bit better haha.
@picuberoot oh it shouldn’t be that different because I have the 8th edition. And ok, thanks! Also, I failed the 2012 one lol, things just didn’t click to me xD
I’m scared of the FRQs
@AnniiT make stuff up for FRQs if you blank and at least get points for interpretations that way, and always write the degrees of freedom and conditions and type of test and copy down everything you see on your calculator just to be safe. You could do te wrong test and still get most of the points.
@stemscholar I’m not good with interpreting data. (Cuss) center, spread, shape, etc.
We use GSOCS: Gaps/shape/outliers/center/spread
Center: mean if given or calculated form data (Average), or median which is the middle number when the numbers are in order least to greatest
Spread: standard deviation or variance, not sure how to explain calculations for those but remember that you cannot add or subtract standard deviations. When combining sets of data, square both standard deviations, (this is the variance), then add, and then take the square root for the new standard deviation
Shape: symmetric/normal/bellshape are most common I’m guessing. It’s uniform if it’s a straight line across. It’s skewed right if the tail is on the right and skewed left is the tail is on the left
Outliers: Q1-1.5(IQR) and Q3-1.5(IQR) and explain which way they affect data.
Put everything in context always. I think that’s all you need to know for that.
I was referring to interpreting things like p value though so if you can’t figure out the p value on frq, make one up and do the right interpretation saying something like “if the null hypothesis is true, then there is a ___% chance of observing a pvalue as extreme as ours” and stuff like that
Can someone explain 2012 #11 and #13 and #15 please?
for part c of 3 on last years free response why wouldn’t the probability be 2/5 * 3/14 *3/13???
"A typical school week consists of the days Monday, Tuesday, Wednesday, Thursday, and Friday. The
principal at High School A believes that the number of absences tends to be greater on Mondays and
Fridays, and there is concern that the school will lose state funding if the attendance count occurs on
a Monday or Friday. If one school day is chosen at random from each of 3 typical school weeks, what
is the probability that none of the 3 days chosen is a Tuesday, Wednesday, or Thursday? "
college board says:
“For any one typical school week, the probability is = 2 0.4 5 that the day selected is not Tuesday, not
Wednesday, or not Thursday. Therefore, because the days are selected independently across the three
weeks, the probability that none of the three days selected would be a Tuesday or Wednesday or
Thursday is .4^3 = 0.064.”
@jamanda for 13 we know that the bunny cannot weigh less than 0 pounds, and the average is 5. If it were normal, about 3 standard deviations on either side of the mean would encompass all of the data, and that would be 3(3) because we have 3 times the std dev, which is 3. However, on the left side, we cannot go that far down because it would be negative and we know that’s impossible. Therefore, the curve is skewed. Hope that helps some
For 11, we know it’s not e because he is applying a treatment, the different detergent. We know it’s not D because matched pairs refers to pairing subjects, not treatments. Wek ow it’s not C because blocking by washing machines would consist of doing a statistical analysis of each washing machine separately, but there’s only one washing machine. It cannot be B because we don’t block by treatment, but rather by population characteristics. It’s A because he randomly flips the coin to assign the treatment. (I thought it was a little iffy to call it completely random, as they are coming in at the same time so may have come in from a similar populations if people use the facilities in groups and those groups have similar characteristics, but nonetheless the other options make less sense)
@blaked There are three weeks total. For each of the three weeks, there is a 2/5 chance of not choosing a Tuesday, wednesday, or thursday. we do (2/5)(2/5)(2/5) = .064. We can do this because the events are independent, and do not affect each other. I’m not sure where you got your numbers on the top of your post. If you don’t get this explanation, feel free to let me know where your numbers came from and I can try to help out from there too.
Does anyone have a list of all sample sizes we use for each test or whatever?
I know for inverse T you use n-2 and linear regression test for slope you use n-1 but I don’t remember what uses n-1 vs just n.
I need help in the 2012 MC #15, #28, #33, and #40. Thanks in advance!
Someone please explain to me the difference between matched pairs and blocking. I need it in actual human words…