(no subject)

[cheekbones3]

Being in a bad mood actually does make me want to post more! Anyone got any crap jobs they need doing? Collation of papers, file renaming that sort of thing. £5 an hour is my rate! In fact, I could do some tutoring actually. Maths, physics, or any science up to 16 year old level, plus geography and related I suppose. I really am that poor. Piss!

Comments

[toffeecat.livejournal.com]

Continuous Probability Distributions
Uniform Distribution
Binomial Distribution

Changing things to Z-Scores


:-/ I am going to see my Math prof. on Weds hopfully he can help me too

[cheekbones3.livejournal.com]

Easy easy easy! Ask me and I can tell :O)

[toffeecat.livejournal.com]

What is the point on Z-scores......what do they measure.......erm, i'll think of more questions tomorrow. Its 1:30am so i am sleepy.....but tomorrow i will ask more

[cheekbones3.livejournal.com]

Z is a simple way to represent probabilities on the normal distribution. It comes from X=rho.Z+mu, where X is a general normal dist., and gives a simple way to measure the probabilty of finding a point within a certain range of this distribution.

Z expresses x (a value from a given distribution) in the number of standard deviations from the mean. A very useful fact is that there is virtually 95% probabilty of finding a value in the range -2<Z<2, since 95% is the most common test for statistical significance. It's easy to grasp the uses of Z with examples - the examples should really easy. Symmetry is also very useful when using Z.

[toffeecat.livejournal.com]

One of the questions is

Suppose that x has a binomial distribution with n=200 and p=0.4

a) Show that the normal approximation to the binomial can appropriately to used to calculate probablities about x.

b)Make continuity corrections for each of the following, and then use the normal aprroximation to the binomial to find each probability
1) P(x=80)
2) P(x<65)
3) p(x>100)

[cheekbones3.livejournal.com]

Erm I never did this, but you can say for a) that the central limit theorem can be used to sum all the independent random variables that make up the binomial r.v., which means that the normal can be used when n is big enough, so mean =n.p and SD=sqroot(n.p.(1-p)).

b)1) Mean=80, SD=6.93, Z=see a table for the mid value of n=200 I think...
2) 0.0125
3) 0.002

I find that I'm actually bloody rusty, so this is helpful work for me as well - that's what I make those to be - could you tell me if I'm right? Lots of practise is the key...