2011 AP Statistics Thread

<p>isnt p the parameter (for the population) and p hat the statistic (for the sample)?</p>

<p>Geometric is when you do something UNTIL you get a success. Binomial is when you calculate the number of successes in a GIVEN number of trials. </p>

<p>Does that make sense?</p>

<p>Can someone explain chi square vs chi square GOF? Like when to use them?</p>

<p>from what i’m reading, parameters and statistics are synonymous to a measurement of a proportion? what i’m confused about is why they’d talk about p hat and then suddenly use p for finding mean and standard deviation for p hat</p>

<p>Here’s an example that might help clear things up in binomial (i’m still confused about cdf!)</p>

<p>A manager notes that there is a .125 probability that any employee will arrive late for work. What is the probability that exactly one person is in a six person department will arrive late?</p>

<p>Note the key word: “exactly” one person</p>

<p>So…binompdf(n,p,x) -> binompdf(6,.125,1)= .385</p>

<p>ok i understand that</p>

<p>but i dont understand #26 and 27 on here: [AP</a> Statistics Practice Exam 2 | Education.com](<a href=“http://www.education.com/study-help/article/ap-statistics-practice-exam-2/]AP”>Education.com | #1 Educational Site for Pre-K to 8th Grade)</p>

<p>like, i dont even know how to start</p>

<p>the “c” stands for cumulative, so you used a PDF there because of the word “exactly”</p>

<p>Can anyone tell me the difference between a sampling proportion and a sample mean?</p>

<p>yea a sampling proportion is x/n but a sample mean is when you’re finding a specific value right? basically, it depends on what you want to find out i think. like if you want to find out the mean length of all fish, you would do a sample and find the sample mean. if you wanted to find out if cell phones or passengers distracted drivers more, you would find the proportions in the sample.</p>

<p>oh and sample mean is x bar, sample proportion is p hat
population mean is mu and pop. proportion is p</p>

<p>Is there an answer key for that test anywhere? (for number 26 & 27)</p>

<p>yes its on page 2</p>

<p>1997 #23, can you help me understand why its c)</p>

<p>A 95% CI of the form p hat (plus or minus) E will be used to obtain an estimate for an unknown population proportion p. If p hat is the sample proportion and E is the margin of error, which of the following is the smallest sample size that will guarantee a margin of erro of at most .08? a) 25 b) 100 c) 175 d) 250 e) 625</p>

<p>The generic is .5/sqrt(n) when the standard population isn’t given</p>

<p>So… critical number for .95 is 1.96:</p>

<p>1.96(.5/sqrt(n)) < .08
Solve for n, you get 150
Since 150 is not an optional number, you pick the closest answer which is 175 because as sample size n increases, margin of error would decrease. If 175 were to be a sample size, margin of error would be less than .08.</p>

<p>Hope that helps!</p>

<p>Someone asked earlier about Type 1 and Type 2 errors.</p>

<p>Type 1: Rejecting Ho when it’s true
Type 2: Accepting Ho when it’s false/ Fail to reject when it’s false
Power: probability of rejecting the false null hypothesis; not committing a Type 2 error/ increasing the sample size and sig. level will help in increasing the power</p>

<p>There’s an example on Barron’s
Suppose the null hypothesis is that all systems are operating satisfactorily with regard to a NASA liftoff. A Type I error would be to delay the liftoff mistakenly thinking that something was malfunctioning when everything was actually OK. A Type II error would be to fail to delay the liftoff mistakenly thinking everything was OK when something was actually malfunctioning. Type II is more severe. Power would be the probability of recognizing a particular malfunction.</p>

<p>@hellogoodbye893: thank you for that explanation! I was about to ask about the same question. </p>

<p>good luck everyone!</p>

<p>(Not sure if anyone still needs it, but on Type 1/2 errors, there’s a quick mnemonic to help remember them: ART is my BFF. Alpha: Reject when actually True; Beta: Fail to reject when actually False.)</p>

<p>What is the difference between chi test for independence and chi GOF?
Sent from my PC36100 using CC App</p>

<p>@AdrianGon
GOF test: used for only ONE variable.(this is the MAJOR difference)
null hypothesis is about the data fitting certain quotas
degrees of freedom: n-1</p>

<p>Test for independence: used for multiple variables
null hypothesis is about the data being independent
alternative hypothesis is data is dependent/association exists
degrees of freedom: (c-1)(r-1)</p>

<p>GOOD LUCK TODAY GUYSS :)</p>

<p>@anotherindiankid</p>

<p>Idk if you still need help with questions 26 and 27 but basically what you have to do is first establish the probability of getting each number of points…</p>

<p>since the probability of baxter getting a shot in is .6 and the possible points that he can get is 0,1,2 the distribution for the probabilities would be…</p>

<p>0: .4 (because he has to just miss the first shot)
1: .24 (.6x.4 because he makes a shot and THEN misses)
2: .36(because he makes BOTH shots)</p>

<p>to find the answer to 26 you just have to see which of the points has the highest probability, which is 0 so c would be the answer.</p>

<p>for 27 you just have to multiple each of the points by their probabilities so it would be…
0(.4)+1(.24)+2(.36)=.96 so the answer would be b.</p>

<p>hope i helped!! good luck today :)</p>

<p>A tree diagram might help with those last two problems, if you like visuals…!</p>

<p>@Adriangon:</p>

<p>For Chi Square GOF test:
Ho: p1=p2=p3=pn
Ha: at least one of these proportions is incorrect
You’ll need to use degrees of freedom: df=n-1</p>

<p>For Chi Square Homogeneity:
Ho: The distributions of x selected are same in all y
Ha: The distribution of x selected are not all the same
If you’re using a calculator, it’d be with matrices</p>

<p>For Chi Square Independence/Association:
Ho: X and Y are independent
Ha: X and Y are dependent</p>

<p>Can someone explain to me that why it’s Ha: p1-p2>0 and not Ha: p1-p2<0 when a hypothesis test for a difference between two means asks “At the 10% significance level is there sufficient evidence that the candidate’s popularity has decreased?”</p>