Okay: on october 24th the second to last slide has the formula for (1-alpha)CI for mu1-mu2 is ... and then he has x1-x2 but sometimes in examples he uses x2-x1. Which ones right. Normally you get the same answer just negative and positive...

11/17/2005 5:57:00 PM

i remember that class when he was using that formula to do an example problem

he started out using x2-x1

people got confused and someone asked why he was changing it

instead of answering, he went off on a tangent and started drawing out this random formula that he eventually told us "has nothing to do with this problem"

then he went back to the example and used x1-x2 instead





I remember this because my friend was visiting from out of town that day and went to class with me and when he was going over the problem she was laughing so hard and then said "i've never seen a teacher explain a problem worse"

11/17/2005 6:21:08 PM

I wrote down most of the stuff he went over in class wed, and i'm putting a few webassign probs down, and some examples from the book, ones that are either good examples of how to do a certain problem or ones that are tricky and could mess me up...that's what's on my sheet.

11/17/2005 11:49:15 PM

man i hate this class you dont know what he is going to put on the test

11/18/2005 1:26:03 AM

heh, nm not knowing what's on the test, half the time we aren't sure of the equation because he does it wrong, says "oops" then just moves on

11/18/2005 3:48:08 AM

comment on the test??
anyone know how to have worked #8, with the SSW and SSB given?
yes or no for #2, about the size(?)

11/18/2005 2:22:56 PM

i put no.


man that sucked

11/18/2005 2:35:58 PM

it was improved, but still too much unexpected stuff

I personally had no idea which regression chart to pick...never remember hearing or seeing anything resembling that question.

Also, the prob. where all we got was the SSW and SSB, you hadta CREATE AN ANOVA CHART and get the F-ration, compare that to pvalue, and determined rejection status. It took me like 6 minutes of "WTF?" to finally figure that out. I would've saved so much time had Reich just put like "Create ANOVA chart" at the end of that problem.

I had no idea about the size, but, I remember him saying that if the sample sd. dev.'s were "close", then it was enough. Plus, on a recent webassign there was a problem w/ one sample sd as .06 and one as .065, and they were "close enough" to do the problem, so I put "yes" for #2.

It'll be interesting to see the scores, as usual.

11/18/2005 2:44:18 PM

i don't think the test was all that terrible. i just used all the formulas in that last powerpoint he gave. for the sample size, i just used if np>5 and n(1-p)>5. and the residual, i think he was just seeing how many people know what residuals are, which is the difference between the slope of the line and the observed point, so if it was +2 off, then it was plotted at 2. kinda like if you rotated the graph so it was hoizontal.

11/18/2005 3:24:40 PM

It wasn't TOO bad...better than the last ones for sure


But yeah, that ANOVA problem....i saved that for last and i FINALLY figured out we needed the F stat to begin and couldn't get it before he said "time's up"

at least I wrote the letter "F" on that problem...maybe I'll get 1 point




But I'm still upset at the fact that this test was still far too advanced for beginning stat students...I plan on being a manager, not a freakin statistician

11/18/2005 3:32:54 PM

I just wrote stuff like ... Hey, this F value is too small to get an a<0.05 and never bothered to actually look it up

I figure if a 65 is a C, I'm happy with an 80

11/18/2005 3:53:44 PM

Quote :

"I plan on being a manager, not a freakin statistician"

So true.

Yeah I just now figured out what he was trying to get across w/ that #5.
Seriously though, that is the MOST OBSCURE AND FAGGISH WAY to see if we knew what residuals were. He could've used one of the 19 methods he showed in class instead of just giving us 4 blank charts to choose from. What a fag.

11/18/2005 3:55:18 PM

someone post when the test scores are up on webassign.

11/18/2005 4:12:58 PM

i think the test was a lot better than the last ones the only problem i had trouble with was # 10 but i did not read that chapter

11/18/2005 4:21:48 PM

So, what was the answer to #5? I guessed D.

OH and exam scores are up - I got 36.5/40 I got there RIGHT before he announced there was 15 minutes left. I had a really really good cheat sheet I guess.

[Edited on November 18, 2005 at 7:28 PM. Reason : test scores are up]

11/18/2005 7:27:29 PM

dude, doesn't he realize that something ain't right when on the first test, he had to curve everyone to make a 75 average

then on the second test, the average is a 65

and on this test, the average is a 60


I sense a pattern

11/18/2005 7:50:30 PM

Meh, I got 33/40...82.5%

Not terrible I guess, but check this: THE CLASS AVERAGE IS A 60. ON THE THIRD TEST OF THE SEMESTER. MAYBE THIS IS A SIGN THAT YOU CAN'T TEACH FOR CRAP REICH

11/18/2005 8:03:40 PM

OMG I FEEL SO STUPID I JUST FIGURED OUT WHERE YALL WERE GETTING THE CLASS AVG INFO FROM

So, when the info says average=66.7 for the exams, does that mean 66.7/102? Because that's like 65% average over all three tests

Like whoa

[Edited on November 18, 2005 at 8:12 PM. Reason : .]

11/18/2005 8:08:06 PM

hehe, don't feel stupid, you made an A++++ after the curve

11/18/2005 8:09:02 PM

this is bull shit how the fuck is he getting these grades i cant fucking believe this shit
a damn 60 for a average is shit

11/18/2005 11:39:32 PM

are these grades with the curve cause if so

11/19/2005 12:26:32 AM

No curve! What the hell is wrong with this guy?

I love on the slides how he put the median grade instead of the average.

11/20/2005 11:07:26 PM

Maybe he is just planning on curveing the final course grades.. either way I am WAY pissed off about my grade. I am sick of feeling like I will never do well in his class....

11/20/2005 11:10:17 PM

SouthPaw -- Can we have some help on the last webassign?
THANKS!

11/27/2005 5:12:05 PM

holy balls.

this douchebag must not understand the concept of dead week

I am amazed that we seriously have one due...I wouldn't have even looked at WebAssign had you not said anything. Should we email him and be like WTF are you doing?

11/27/2005 5:38:41 PM

I think there is a lab too but someone double check that for me

11/27/2005 6:25:52 PM

yeah webassign, stupid. probably is a lab too.

11/27/2005 7:14:32 PM

webassign+lab (during dead week) = new asshole for reich on evaluations

11/27/2005 7:56:20 PM

After calculating the regression for a spreadsheet, how do I find the answer to questions such as "According to the regression, how many additional years of life can be expected if an additional 10% of 1-year olds have been immunized for polio?"
BTW I have all the correct figures in my regression, just not sure where to look for these...

11/28/2005 11:52:23 AM

Just a headsup: he said in class today that he'll drop your 2 lowest WebAssigns. Thus, if you have 7 WebAssigns with a satisfying grade (to you), you don't even hafta attempt this last one.

11/28/2005 2:10:00 PM

i didnt understand what he was saying about curving grades. like he said they look on par with other classes so he wont curve final grades?

11/28/2005 2:47:46 PM

I think he is trying to tell us, in not so many words, that he will be curving our grades a lot. He said a 65 will be entered as a solid C so 60 is prob a C- and 75 is prob a B I am guessing an A is anywhere from 85 up

but I do not think he will be curving the final exam grades

[Edited on November 28, 2005 at 4:39 PM. Reason : exams]

11/28/2005 4:39:15 PM

Quote :

"The final exam will be open book, open notes, anything you want other than a laptop."

11/29/2005 2:18:29 PM

^where did you get this? i didnt hear him say it in class

[Edited on November 29, 2005 at 2:22 PM. Reason : ks]

11/29/2005 2:21:44 PM

He emailed us today!!!! by the way I am seeing my tutor tonight so I will try to post what I can figure out about the webassign and lab

11/29/2005 3:00:54 PM

ah right on

11/29/2005 3:32:39 PM

any help you can give on the hw would be greatly appreciated thanks

11/29/2005 7:12:21 PM

Okay.... Get ready to use Excel

problem 1

you can literally draw it out by hand if you want but I recommend putting the data into Excel, I put the Y's in as A and the X's in as B. Then highlight everything you entered in and go to the picture of the bar graph, click that and pick that you want a XY (scatter). Make sure it’s clicked onto columns then click "finish"... And there is your scatter plot.
(a) its positive, because of the way the scatter is flowing
(b)To find the correlation, go to "tools", "data analysis" and then pick "correlation"... Mine didn't have it on there already so I had to go to "add on" and actually kind of install it. I am NOT an excel wiz and neither was my tutor but this works.... For "imput range" you want to click on the little picture next to the space that looks like an excel chart then high light the A and B columns. And click the picture again next to the space and it will take you to a "new worksheet"... The number in there is the answer.
(c) Ok for this one you are going to use the formula on page 475 of our books, "The Test Statistic for Correlation"... your "r" is from part b, and n 20for us, that was given. Solve and put into webass.
(d) then if you want you can actually solve on the t chart using the number from c and your df, but if you don’t then you can guess like I did, and you end up rejecting because your answer in c will be higher then your t test number.
(e) now this one, my tutor explained to me but I don’t think I could explain it as well, so I picked that it increases as there is an increase, but there are 2 answers and 3 tries...

problem 2

(a) Use this equation, = 200 + 150x and put your red number in for x, Volia, there is your answer
(b) Use the equation on page 504 in the textbook “Confidence Interval for E(y)given Xp”… fill in part a for the t value, my t alpha/2 was 2.1199 because my df was 16 but you should know how to find that by now (hint: use the t chart), the SE number is actually the sq rt of MSE like if we using an ANOVA table, so mine was 25/16 (SSE/df) so I got 1.56 then sq root of that gives you 1.249 (? that’s your SE) then fill in the rest of the formula, your red number is Xp, X bar is given, and the sum of blah blah is given too. … Solve, and that’s your CI. I had answers like 8600.6249 and 8599.3751.
(c) use the formula on page 505 of the text book, the only difference is now you are adding 1 inside the big sq root.

I am going to eat some dinner and I will finish #3 when I come back, its not bad and wont take much time.. its just Excel too but if you dont know how to use it then its tricky. Hope these help, if it doesn't I can try to clarify when I come back.

11/29/2005 8:02:22 PM

anyone work on the lab yet?

11/29/2005 9:31:22 PM

Sorry that took so long, I got caught up in Sex and the City and was typing really slow

problem 3

Open up his excel document and the first thing I did was I copied the information I needed into another spread sheet. You will need the “horsepower” and “price as tested” and I copied the names of the cars. Now you should highlight the price and the horse power then make a scatter plot.

(a) First you can see that it is a positive linear regression, so “1”
(b) Find the correlation the same way you did above in number one for this answer

Go to “Tools”, “Data Analysis”, and find “Regression”... then once that box comes up you want to check the box for “Residuals” and “Residual Plot” and then at the top of that box you want to make the “Y” your price column, and make “X” your horse power column. So you would pick these the same way as before like clicking on the picture box then highlighting the column you want... etc…
(c) So for this part about line 17 ish, you want to find the “Coefficient Intercept”
(d) Right below that is the “Coefficient X Variable” that’s what you want
(e) Find the ANOVA table on this page, and find “Regression” “Significance F” but he want you to use all 9 decimal places so its….000000764 (I think it’s the same for everyone)
(f) You can go back and forth to find which column matches up with the ford Taurus and I think it’s #2, you use that number under residual for the answer, but I stretched out that column and got more decimal places…
(g) Now at the top of that page is a graph that I stretched out so it was giant and then I held my mouse over the highest point and then you use the biggest number in its set of coordinates. And that’s your answer…


Let me know if this helped at all, I wish I could have gotten to the lab, if anyone has that worked out, even fake answers would be awesome… Thanks

11/29/2005 10:52:24 PM

We want you back southpaw.....some people aren't as technologically advanced in excel and its conveniently not working ....we know you know how to do it on the calculator!!! so hook a brotha up. word to your motha. star trek 4 life

11/29/2005 11:38:48 PM

geeeez is anyone out there in this galaxy? i mean who doesn't check wolf web every 5 minutes. geeeez its not like you're in a black hole....

11/29/2005 11:51:50 PM

lol, SouthPaw is taking this WebAssign off and not even attempting it...my mood may change, but since he's dropping it I have little incentive to even try

11/30/2005 1:08:25 AM

Packangel thanks for the help.
Excel makes this shit so easy it took all of 15 minutes. I kinda needed a 100 on this since i missed 1 webassign some weeks ago.
Who's your tutor btw?
Southpaw, once again i think i speak on behalf of everyone when i thank you for all the help throughout the semester. You kinda took a bit of the sting outta this otherwise painful class and Mr Briian "retard" Reich.

11/30/2005 7:37:32 AM

My tutor is Kelci Miclaus, She is really nice and can normally meet with me most times when I need her. She has been a huge help!

11/30/2005 9:27:05 AM

yes, i'd like to thank you again southpaw for all the help you've given throughout the semester. Without my 100% webassign scores I'd most definatively be failing this class. If I ever see you at a bar, you can bet your ass you and your roommate will be getting drunk on my tab

11/30/2005 11:07:41 AM

southpaw, thanks for all your help... I would most def. not be able to pass without your great explanations!

11/30/2005 11:14:30 AM

anyone go to class today?

11/30/2005 2:42:23 PM

Thanks everyone!

I went to class today, he basically just went over more regression stuff (new material). However, it was nearly all Excel stuff that can't be tested. No mention of an all-out review on Friday or anything exam-related.

11/30/2005 2:50:26 PM

we also filled out evaluations today so that was fun

11/30/2005 3:04:54 PM