A: -6.4 6.6 12.9 3.2 B: -6.2 19 -7.7 4 It asks for the covariance…I plugged the numbers into their respective data points in my calculator and got 10.37…wrong answer. Correct answer was 13.82 The explanation says, when done out by hand, you would use n-1 in the denominator when dividing the summation of all the covariances…i thought this was only done for samples? Help please, thanks.
Wait, how do you calculate covariance on the calculator? And yes, n-1 is only done for samples…can you post the text before these numbers, it may explain why its not a sample.
harley…your numbers must be off somewhere. i did it on EXCEL just now for you to see. i’ll send it over if you want it. e-mail me at: HARVEY.CHAN214 @ GMAIL *DOT* COM
If I don’t know how to do stuff like this, and std dev, and variance, etc. in my calculator, is it conceivable to learn how to do it in the next hour or two? Is it worth it?
I did it and got r=.13945, Sx = 8.05, Sy = 12.302. (I used the sample stdn dev, not the population) .13945 = x/(99.032) x = 13.81
billy22g Wrote: ------------------------------------------------------- > If I don’t know how to do stuff like this, and std > dev, and variance, etc. in my calculator, is it > conceivable to learn how to do it in the next hour > or two? Is it worth it? For STD DEV and variance, press 2nd+7 on your TI. enter in your variables as X0, x1, etc. Then press 2nd+8. Scroll down to std dev measurement. Could someone please post how to do covariance on calculator?
On the ti 2nd Data (7 key) enter x and y varianbles Quit 2nd Stat (8 key) use arrows to find r (correlation) and Sx is sample stnd dev of x, the (standard deviation greek letter) is the population stnd dev. Saves a lot of time. easy to learn.
great post june. all this time i’ve been doing the mean and mean differences by hand ===> entering those figures into the data point slots ===> dividing XY by 3. you saved me a lot of time with this new method.
I took it from another poster on this board…cpk maybe.
well thanks to both you and cpk then. any fast ways to do #115 in the mock exam morning session?
kinda…the stat function gives each entry an equal probability, so when you have unequal weights or probabilities like Q15, you can do the following x1 & 2 = -10 x3 through 5 = -5 x6 through 9 = 10 x10 = 25 get stnd dev of the asset and square it. = 121 Not sure its faster though.
you so lost me… but definitely thanks for the other one cause my head was spinning going back and forth from paper to calculator trying to get it.
What I meant by the fact that I used my calculator was that I was able to get the std devs of a and b and also the correlation coeff. so i just multiplied them out to solve for cov.
This question was 117# in the mock morning exam. It doesn’t have any texts. Does anyone know what would indicate that it should be computed as a sample?
Sorry…for Q 115: look at the probability part of the question. since they arent all equal, you need to “trick the calculator” becuase the calc thinks every X and Y you enter into the sheet have an equal probability. To trick the calc, a 20% probability is just two 10% probabilities, so enter the x value 2 times. We’ll use 10 total inputs, so 2/10 is 20%. Same for a 40% probability…just enter the value 4 times. apply this to the full set and you end up with 10 total x values with probabilities equal to the number of times you used the value…
That’s so awesome. Thanks so much for that guys. Helps a ton. What do the other variables mean? (a, b, Sigma(X), Sigma(X^2), etc.) No easy way to get around hypothesis testing with similar methods, is there?
june2009 Wrote: ------------------------------------------------------- > I did it and got r=.13945, Sx = 8.05, Sy = 12.302. > (I used the sample stdn dev, not the population) > > .13945 = x/(99.032) > > x = 13.81 Got the same thing. What is this 99.032 are you just calculating it out or something
^^^ 99.032 is product of 8.05 and 12.02…the stnd devs
^^ billy the other numbers are for linear relationships, regression and intercept. this function assumes x (independent) and y (dependent) variables. it is immaterial for what we will be tested on.
got it guys. thanks for everything. GOOD LUCK TOMORROW