The Student Room Group

AQA A2 Mathematics MS2B Statistics 2B - 21st June 2016

Scroll to see replies

Reply 20
Original post by student0042
I've taken that one and M3. I thought FP3 was strange but I could still manage it, but M3 was just much harder than usual. So, I'll have to be prepared for whatever they throw at us.


Yeah I've heard about M3, a lot of people found it extremely difficult
I'm finding it difficult to decide which distribution to use.

Does any one have any tips on how to recognise if you're using Poisson, Binomial or Normal ?
Haven't had much time to prepare for this one, Do you think an A is doable if I work hard over the next 3 days? :frown:
Reply 23
Original post by Krithika256
Haven't had much time to prepare for this one, Do you think an A is doable if I work hard over the next 3 days? :frown:


Yes! at least 2-3 questions in this exam are pretty much exactly the same with a slight difference in distributions you use. work your ass off and you'll be fiiine
Reply 24
Original post by Roxanne18
I'm finding it difficult to decide which distribution to use.

Does any one have any tips on how to recognise if you're using Poisson, Binomial or Normal ?


Firstly, it's important to understand that the Poisson and Binomial distributions are discrete distributions, whereas the Normal distribution is a continuous distribution.

I have listed a few things to look out for below:

Poisson

We are often told that the discrete random variable in the question can be modelled by a Poisson distribution.

If not, phrases like 'constant average rate' are the giveaway sign.

Binomial

Binomial questions don't seem to come up a lot in S2, but they are sometimes thrown in amongst Poisson questions.

There must be a fixed number of independent trials for the binomial distribution to be applicable. The trials must also have 2 outcomes, success or failure, and the probability of success must be constant for each trial.

Normal

In S2, the standard normal distribution is mainly used to calculate confidence intervals and carry out hypothesis tests. Be careful here though, as depending on the information given in the question, we may have to use the Student's t-distribution instead.

Hope that helps :smile:
(edited 7 years ago)
Original post by sam_97
Firstly, it's important to understand that the Poisson and Binomial distributions are discrete distributions, whereas the Normal distribution is a continuous distribution.

I have listed a few things to look out for below.

Poisson

The examiner often tells us that the discrete random variable in the question can be modelled by a Poisson distribution.

If not, look for phrases like 'constant average rate'.

Binomial

Binomial questions don't come up a lot as they are mainly covered in S1, but they are sometimes thrown in amongst Poisson questions.

There must be a fixed number of independent trials for the binomial distribution to be applicable. The trials must have 2 outcomes, success or failure, and the probability of success must be constant for each trial.

Normal

In S2, the standard normal distribution is mainly used to calculate confidence intervals and carry out hypothesis tests. Be careful here though, as depending on the information given in the question, we often have to use the Student's t-distribution instead.

Hope that helps :smile:


Thank you so much!!! (:
Reply 26
can someone help me for wording answers to questions about whether someone claim is supported or not and how to get the correct wording for the explanation mark after you reject or accept null. I always word it the wrong way to them. If there's a set method for writing it properly in different scenarios someone please tell me so I can learn it.
Reply 27
Original post by Roxanne18
Thank you so much!!! (:


No problem :smile:
Original post by KB_97
can someone help me for wording answers to questions about whether someone claim is supported or not and how to get the correct wording for the explanation mark after you reject or accept null. I always word it the wrong way to them. If there's a set method for writing it properly in different scenarios someone please tell me so I can learn it.


I kinda struggled with this. You need to look what the questions asking, if it's a claim you either have evidence or no evidence to support the claim, you don't conclude it's incorrect you just don't have evidence to support it. I do agree it can be tricky with some questions though
Hey :smile: I have a stats question which seems really simple but is bugging me. (its question 6 (c) form June 2011)
Screen Shot 2016-06-19 at 13.33.49.png

F(q3)=0.75
so Integrating 0.25(5-2x) between q3 and 1 should = 0.75 right?
In the mark scheme it says

0.5 + [integral of 0.25(5-2x) between q3 and 1] = 0.75

and not [integral of 0.25(5-2x) between q3 and 1] = 0.75
(edited 7 years ago)
Original post by salome12
Hey :smile: I have a stats question which seems really simple but is bugging me. (its question 6 (c) form June 2011)
Screen Shot 2016-06-19 at 13.33.49.png

F(q3)=0.75
so Integrating 0.25(5-2x) between q3 and 1 should = 0.75 right?
In the mark scheme it says

0.5 + [integral of 0.25(5-2x) between q3 and 1] = 0.75

and not [integral of 0.25(5-2x) between q3 and 1] = 0.75


In part b) you showed that 1 is the median. (which is at the end of the first function)
You can't just integrate the second half of the function and equal to 0.75.
Q3 is only 0.75 for the whole function, so you need to add the first bit on aswell, if that makes sense ?
That's where the 0.5 (median) + integral comes from
Took me ages to get my head round this question, sorry if it doesnt make sense!
Original post by Roxanne18
In part b) you showed that 1 is the median. (which is at the end of the first function)
You can't just integrate the second half of the function and equal to 0.75.
Q3 is only 0.75 for the whole function, so you need to add the first bit on aswell, if that makes sense ?
That's where the 0.5 (median) + integral comes from
Took me ages to get my head round this question, sorry if it doesnt make sense!


oooohhhhh I get that, so the integration of the second function was 1/4 of whole function , thank you so much !
Original post by salome12
oooohhhhh I get that, so the integration of the second function was 1/4 of whole function , thank you so much !


no problem (:
Why do you have to use 0.975 instead of 0.95 and when do u know when you must do that because sometimes the mark scheme says use 0.95
Does it occur for other values as well
Feeling decent about this, have M1 at the same time so I sacrificed M1 revision for S2. Think I need about 20 UMS in M1 to get an A overall so hoping it was worth it so my S2 can go to Further Maths
Original post by yoda123
Why do you have to use 0.975 instead of 0.95 and when do u know when you must do that because sometimes the mark scheme says use 0.95
Does it occur for other values as well


So for ex. doing a 5% significance level Hypothesis test

If it is a Directional (one-tailed) hypothesis test (i.e. H0 u=50 , H1 u<50 or u>50 ) you use
p(α) = 0.95

if it is a non-directional (two-tailed) hypothesis test (i.e. H0 u= 50 , H1≠50 )
you us
p(α/2)= 0.975

From June 2014-
"Carry out a hypothesis test, at the 10% significance level, to examine whether themean time for an annual service carried out by Gary is 20 minutes."

This is a non-directional (two -tailed ) hypothesis test and you would use
H0: u=20 H1: u≠20
and p(α/2)=0.95

"Rajul suspects that Gary may be taking longer than 20 minutes on average to carryout an annual service. Find the highest value of the sample mean which would not support Rajul’s suspicionat the 5% significance level. "

This is a directional hyp. test so you would use
H0: u=20 H1: u>20
and p(α)=0.95


With contingency table (X^2 distribution), you always use p(α)
ie
"Carry out a x2-test at the 10% significance level to investigate whether there is an association between country and gender of recruits."

so p(α)=0.9
(edited 7 years ago)
http://filestore.aqa.org.uk/subjects/AQA-MS2B-W-MS-JUN14.PDF

6(a)

can anyone help with the E mark?

the question says to test if the mean time is 20 minutes and we accepted h0. So I concluded the mean time is 20 minutes but they used that there's no evidence to say it isn't 20 minutes. Am I incorrect? Report also mentions my answer but I don't really get what they're getting at (http://filestore.aqa.org.uk/subjects/AQA-MS2B-WRE-JUN14.PDF)
Original post by Rabadon
http://filestore.aqa.org.uk/subjects/AQA-MS2B-W-MS-JUN14.PDF

6(a)

can anyone help with the E mark?

the question says to test if the mean time is 20 minutes and we accepted h0. So I concluded the mean time is 20 minutes but they used that there's no evidence to say it isn't 20 minutes. Am I incorrect? Report also mentions my answer but I don't really get what they're getting at (http://filestore.aqa.org.uk/subjects/AQA-MS2B-WRE-JUN14.PDF)


Basically whenever we 'accept H0', we never prove u=20 (or whatever it is), it just means that we show there isn't enough evidence to disprove u=20 .

When we 'reject H0', we say 'there is statistically significant/enough evidence' to disprove H0.

like if you carried out a 5% significance test and you reject H0, that means "there is enough evidence at a 5% level to disprove H0".
If you then did a 1% significance test and accept H0, that means "there is not enough evidence at a 1% level to disprove H0".
Original post by salome12
Basically whenever we 'accept H0', we never prove u=20 (or whatever it is), it just means that we show there isn't enough evidence to disprove u=20 .

When we 'reject H0', we say 'there is statistically significant/enough evidence' to disprove H0.

like if you carried out a 5% significance test and you reject H0, that means "there is enough evidence at a 5% level to disprove H0".
If you then did a 1% significance test and accept H0, that means "there is not enough evidence at a 1% level to disprove H0".


Would we be marked wrong if we said " Evidence suggests that the mean time is 20 " Something along them lines ?
Original post by Roxanne18
Would we be marked wrong if we said " Evidence suggests that the mean time is 20 " Something along them lines ?


yeah that's fine, just as long as you don't say evidence 'proves' it :smile:

Quick Reply