Leaving Certificate: Higher Level Maths Discussion

Midlife Crashes · 08-06-2014 10:00pm

Is there anything to have learned off besides trig proofs theorems and constructions ??

allyb17 · 08-06-2014 10:01pm

Ok I'm after hearing that margin of error has changed to like 1.96 root of the standard deviation over n, is this true? I always thought it was 1/root n ?

Also could someone tell me how to find the median from a histogram? Is there a certain formula?

Thanks in advance

Isitthough? · 08-06-2014 10:02pm

Midlife Crashes wrote: »

Is there anything to have learned off besides trig proofs theorems and constructions ??

Well theory on statistics like What types of sampling. that's it I think

Ompala · 08-06-2014 10:02pm

aleatorio wrote: »

This is from the Examcraft mock and I think it's the Central Limit theory?

'400 samples of size 10 are taken from a production line producing jars of gravy. The mean weight of a jar is 200g with a standard deviation of 1.5g. Calculate the limits between which you would expect 95% of the sample mean to lie'

I know the formula for the mean of the sample is the SD divided by the sample size, but is the sample size 10 or 400(10)?

And then it would just be 200+/- 2 standard deviations.. Yeah....?

I dunno if this is ever mentioned in LC, but the reason 400 is important is that if the number of samples is greater than 30, then the distribution will be a normal distribution.

If less than 30, you have to use a different set of tables.

Aspiring wrote: »

Didn't know you could do standard deviation on your calculator till now... It's pretty much the same as finding correlation coefficient except easier.

Be very careful with this.... do you use a button with Sx written on it to calculate the standard deviation?

robman60 · 08-06-2014 10:05pm

I'm sure this has been asked previously, but are there any solutions available for the 2014 sample papers? I'm flummoxed with questions 4 and 5 but I haven't found anything yet. I'll pay a few euro if necessary because if that question on the circle comes up I won't have a clue.

Isitthough? · 08-06-2014 10:05pm

allyb17 wrote: »

Ok I'm after hearing that margin of error has changed to like 1.96 root of the standard deviation over n, is this true? I always thought it was 1/root n ?

Also could someone tell me how to find the median from a histogram? Is there a certain formula?

Thanks in advance

I think I did a question like that where it was like a set of races and they gave you 3 Histograms..There wasnt like a formula you just need to know if is positive/negative skew then you'll know that either mean>median>mode or vice versa and use that knowledge to make an educated guess. If Im wrong someone tell me cause thats just me if theres a actual way Id love to hear it

lostatsea · 08-06-2014 10:11pm

robman60 wrote: »

I'm sure this has been asked previously, but are there any solutions available for the 2014 sample papers? I'm flummoxed with questions 4 and 5 but I haven't found anything yet. I'll pay a few euro if necessary because if that question on the circle comes up I won't have a clue.

http://www.studentxpress.ie/HLProjectMaths/index.html

aleatorio · 08-06-2014 10:13pm

Ompala wrote: »

I dunno if this is ever mentioned in LC, but the reason 400 is important is that if the number of samples is greater than 30, then the distribution will be a normal distribution.

If less than 30, you have to use a different set of tables.

Be very careful with this.... do you use a button with Sx written on it to calculate the standard deviation?

Oh really? I was wondering why they said the 400 part

thanks!

Ompala · 08-06-2014 10:18pm

aleatorio wrote: »

Oh really? I was wondering why they said the 400 part thanks!

I could go on... but I don't want to be remembered as the guy who confused the hell out of you the night before your maths exam :pac:

FatRat · 08-06-2014 10:20pm

They haven't asked for a proof of a trig identity yet on project maths... Perhaps this could be the year! Hopefully anyway because I spent so long learning them...

allyb17 · 08-06-2014 10:21pm

Isitthough? wrote: »

I think I did a question like that where it was like a set of races and they gave you 3 Histograms..There wasnt like a formula you just need to know if is positive/negative skew then you'll know that either mean>median>mode or vice versa and use that knowledge to make an educated guess. If Im wrong someone tell me cause thats just me if theres a actual way Id love to hear it

I thought so! Thanks a mill, what do you know as the margin of error formula though I don't know which to use

FatRat · 08-06-2014 10:21pm

Ompala wrote: »

I could go on... but I don't want to be remembered as the guy who confused the hell out of you the night before your maths exam :pac:

This isn't on the LC course anyway. We only use z-scores.

aleatorio · 08-06-2014 10:21pm

Ompala wrote: »

I could go on... but I don't want to be remembered as the guy who confused the hell out of you the night before your maths exam :pac:

Ah sure go on, I like maths I won't be too confused I'm sure :pac:

robman60 · 08-06-2014 10:24pm

lostatsea wrote: »

http://www.studentxpress.ie/HLProjectMaths/index.html

I love you. Thanks so much.

TooMuchStudy · 08-06-2014 10:27pm

Really really hope its a nice paper tomorrow

Ompala · 08-06-2014 10:38pm

FatRat wrote: »

This isn't on the LC course anyway. We only use z-scores.

Look away now! :P

aleatorio wrote: »

Ah sure go on, I like maths I won't be too confused I'm sure :pac:

Basically you use a different distribution known as the t distribution when the number of samples is less than 30. It allows greater variation in your data when conducting hypostesis tests or confidence intervals, as well as if you had to use linear regression. It can also follow that you can pool sample sizes etc with it. The larger the sample size gets the distribution converges towards the normal distribution.

Now I'm only in Engineering so maybe some Mathematicians could correct me, but thats the bones of it. But it won't be on the paper tomorrow so no need to worry about it

Aspiring · 08-06-2014 10:40pm

Separate line and circle questions would be lovely :pac: 50 marks in the bag straight off.

FatRat · 08-06-2014 11:07pm

Aspiring wrote: »

Separate line and circle questions would be lovely :pac: 50 marks in the bag straight off.

The more geometry the better for me too! I hope the test is riddled with it, along with a construction and a proof!

aleatorio · 08-06-2014 11:12pm

Ompala wrote: »

Look away now! :P

Basically you use a different distribution known as the t distribution when the number of samples is less than 30. It allows greater variation in your data when conducting hypostesis tests or confidence intervals, as well as if you had to use linear regression. It can also follow that you can pool sample sizes etc with it. The larger the sample size gets the distribution converges towards the normal distribution.

Now I'm only in Engineering so maybe some Mathematicians could correct me, but thats the bones of it. But it won't be on the paper tomorrow so no need to worry about it

Ah that doesn't seem too nasty, I don't think you've compromised my mind for tomorrow :pac:

Least now I'll remember that if they ask something about sample sizes :pac:

MegGustaa · 08-06-2014 11:22pm

aleatorio wrote: »

Ah that doesn't seem too nasty, I don't think you've compromised my mind for tomorrow :pac:

Least now I'll remember that if they ask something about sample sizes :pac:

Also worth noting that if the sample size itself is too small (less than 10ish) and the original population was not normally distributed, the CLT can't be used - that does appear to be on the course!