Edexcel S2 Exam Paper January 2011 14/01/11

Scroll to see replies

The variable T Distributed by N(8, 4), why does P(T = 5) = 0?
Isn't it meant to be P(4.5 < T < 5.5)?

Because with a normal distribution, the probability that it occurs at a single point is always zero, you can only use a normal distribution when investigating the probability a value is above, below, between two certain points, i.e. an inequality. When you change the probability, like from P(T=5) to P(4.5 < T < 5.5), you're thinking about when you change the type of distribution, or something like that, it's been a while.

Reply 41

anshul95

Original post by jit987

The variable T Distributed by N(8, 4), why does P(T = 5) = 0?
Isn't it meant to be P(4.5 < T < 5.5)?

This only applies if you are approximating a discrete distribution like binomial or poisson onto a continuous distribution like the normal distribution. In this question you aren't approximating anything. The question STATES that T follows a normal distribution. As the normal distribution is continuous the probability that it takes any particular value is zero. Think about it: the probability is represented by the area under the graph for the pdf of the normal distribution. At one particular value this is the area of a line which is zero.

Reply 42

anshul95

Original post by Noble.

it clearly has :tongue:

Reply 43

Noble.

Original post by anshul95

it clearly has :tongue:

What's wrong in that statement?

Reply 44

anshul95

Original post by Noble.

What's wrong in that statement?

I was being sarcastic - now that I read back I wasn't being clear enough sorry. You actually explained it very well for someone who potentially hasn't done this for a year.

Reply 45

Noble.

Original post by anshul95

I was being sarcastic - now that I read back I wasn't being clear enough sorry. You actually explained it very well for someone who potentially hasn't done this for a year.

Haha, I was interested to see how much of it was nonsense after not doing it for a while. I have a pretty good memory, so tend not to forget maths, even if it is useless.

Reply 46

anshul95

Original post by Noble.

Haha, I was interested to see how much of it was nonsense after not doing it for a while. I have a pretty good memory, so tend not to forget maths, even if it is useless.

Thats what people say about me as well. Its probably because we enjoy maths so much its not like doing work - well thats how I feel when I'm doing something relating to maths.

Reply 47

jit987

Original post by Noble.

...

Original post by anshul95

...

Thanks! Well explained :biggrin:

Reply 48

gupta21

how do you know when too use the continuity correction? and if so when do you -0.5 and +0.5??
thanks

Reply 49

twiddle-doo

Original post by gupta21

how do you know when too use the continuity correction? and if so when do you -0.5 and +0.5??
thanks

you'd use the continuity correction when you're approximating a poisson or binomial distribution with a normal distribution.

Example:
when the discrete distribution is
P(X<17)
continuous would be
P(X<16.5)

BUT

when the discrete distribution is
P(X>17)
continuous would be
P(X>17.5)

ALSO: because for a normal distribution you cannot approximate 'equals to' so,
when the discrete distribution is
P(X?17) [The '?' should be 'less than/equal to' everytime I save it changes to '?' sorry about that :s-smilie:

)
continuous would be
P(X<16.5)

So when you are calculating the probability for X 'LESS THAN', you'd use the lower bound of X
and
when you are calculating the probability for X 'GREATER THAN', you'd use the upper bound of X

Hope that helps :smile:

Good luck!

(edited 13 years ago)

Reply 50

gupta21

thank you so much 'twiddle-doo'
i finally get it :smile:

Reply 51

twiddle-doo

Original post by gupta21

thank you so much 'twiddle-doo'
i finally get it :smile:

no probs

Reply 52

jit987

Hope Sampling Distribution question pops up :biggrin:

Reply 53

twiddle-doo

Original post by jit987

Hope Sampling Distribution question pops up :biggrin:

I just hope no 'define blah blah' comes up :s-smilie:

Reply 54

Slowbro93

Original post by jit987

Hope Sampling Distribution question pops up :biggrin:

Same! Although I still cannot remember how to define it in a statement, aghhh :frown:

Original post by Noble.

....

This makes sense, thanks!!!! :biggrin:

Reply 55

Foxer27

For the continuity correction

when the discrete is P(x>17)
isnt the continuous P(x>16.5) the lower band?

Reply 56

twiddle-doo

Original post by Foxer27

For the continuity correction

when the discrete is P(x>17)
isnt the continuous P(x>16.5) the lower band?

yeah i had the same question when i was doing this normal distribution but my teacher said no for P(X>17) would be P(X>17.5)
and the reason being for this is because we want values above 17 and 16.5 can be rounded DOWN to 16 than 17

hope that helps :smile:

Reply 57

cazzy-joe

Original post by Foxer27

For the continuity correction

when the discrete is P(x>17)
isnt the continuous P(x>16.5) the lower band?

It is

P(x>17)

not

P(x \geq 17)

If it would have been

P(x \geq 17)

, then it would be

P(x \geq 16.5)

If I'm wrong please correct me :biggrin:

(edited 13 years ago)

Reply 58

Geniusmiss

Original post by jit987

If you can handle those from 01-10 and solomon's very well, then there shouldn't be any problem. How did you find the solomon's?

solomons were ok...i hope nothing like P(4<X<5) comes because i alwys mess up those....any particular tricks to learn how to solve those??

Reply 59

jit987

Original post by Geniusmiss

solomons were ok...i hope nothing like P(4<X<5) comes because i alwys mess up those....any particular tricks to learn how to solve those??

Is that for Uniform distribution questions, like in some interval (0,20) then work out P(4<x<5). Or are you talking about poisson/binomial/cdf/pdf?
For a PDF question if the questions says P(x=5) is it always 0 or is that only for CDF?