The Edexcel S2 (24/05/12 - AM) Revision Thread

Also, for a sample you can test to destruction. An example is testing the lifetime of batteries, if you test all of the batteries using a census, then you would have no batteries left; using a sample means you only test some of the batteries so you will still have some left.

Do any of you guys have the Jan12 question paper and mark scheme?

Adding to raheem's definitions, I've taken the definitions from past papers and bolded the words that you must have to get the marks.

Thanks for this. It's very useful.
Where are Raheems definitions?

Thanks for this. It's very useful.
Where are Raheems definitions?

the boundaries were bloody high for this paper..i was doin the jan06 paper this morning, and it was much easier and also had way lower boundaries(58/75)..edexcel takes the p***

Eh, boundaries should be lower than that. More people taking it in June = more variation in results = lower boundaries. Generally speaking

Eh, boundaries should be lower than that. More people taking it in June = more variation in results = lower boundaries. Generally speaking

Adding to raheem's definitions, I've taken the definitions from past papers and bolded the words that you must have to get the marks.

Critical Region: The range of values of the test statistic that would lead you to reject the null hypothesis.
Significance Level: The probability of incorrectly rejecting the null hypothesis.
Sampling Frame: A list of all the members of a population.
Population: Collection of all items.
Statistic: A random variable that is a function of the sample values which contains no unknown parameters.
Sampling Distribution: All possible samples of the statistic and their associated probabilities.
Census: Where every member of the population is investigated.
Sampling Unit: An individual member of the sampling frame.
Sample: A subset of a population.
Hypothesis Test: A mathematical procedure to examine a value of a population parameter proposed by the null hypothesis compared with an alternative hypothesis.

Assumptions for Binomial:
Occurrences are independent.
The probability of each outcome remains constant.
Two distinguishable outcomes.
Fixed number of trials.

Assumptions for Poisson:
Events occur singly.
The rate remains constant.
Events are independent.

Binomial approximated to Poisson:
n is large.
p is small.
Mean is near to the variance.

Binomial approximated to Normal:
n is large.
p is close to 0.5.

Poisson approximated to Normal:
Lambda is large.

A sample instead of a census
Saves time.
Cheaper.
Easier.

A census instead of a sample
Known population.
Easily accessible.
Total accuracy.

Continuity correction
Normal is continuous, whereas poisson and binomial are discrete.

Hey...

Can someone please explain how to do this question?

4a) set no. of trials, only two possible outcomes, events are independent, probability is constant

4b)i) Die is biased so the probability of 5 coming up will 2/7, and the probability of all the others will be 1/7 - adding up to 7/7, of course.

First five occurs on the sixth throw:

So, (1/7)^5 for the first five throws where 5 DOESN'T come up. And then times that by (2/7) for the sixth throw when 5 DOES come up.

ii) 8C3 x (2/7)^3 x (5/7)^5

Hope that helped Let me know if they're correct

For hypothesis testing, say you have cars passing traffic x~b(50, 0.2)

and there is 7 cars that pass a traffic lights

how do you know if to consider

p(x>=7) or p(x<=7) ?

I know stupid question, but I don't understand how the probability of a 5 coming up is 2/7 and why not 2/6?

Edit: is it because all the other faces are likely to appear? and if the probability of a number 5 appearing was 1/3 then there wouldn't be an equal probability of the other faces appearing?