The Student Room Group

Edexcel S3 - Wednesday 25th May AM 2016

Scroll to see replies

With contingency tables, what do we do if we get an expected frequency less than 5, I've never seen one, but what if?

Also in the conclusions for those questions, do we usually say "not enough evidence for association" or "they are not independent", because I'm getting mixed messages from the mark schemes. Sometimes they seem to be sure like "Money has NO relation to happiness" and sometimes they're like "Oh, I guess we don't have enough evidence of a relationship between money and happiness". I'm like, bitch, MAKE! UP! YOUR! MIND!!
Original post by Euclidean
I agree, just gotta work on my Chapter 5. I was caught out by a question in a solomon paper, basically they gave a table but they wanted you to modify the table and then calculate the coefficient. Quite clever question actually


was it about temperature and finishing position got caught out by it. I ranked one in the wrong order, taught me not too rush things, be wary.

Are solomon S3 easier than edexel papers? Difficulty about the same I think.
Original post by gagafacea1
With contingency tables, what do we do if we get an expected frequency less than 5, I've never seen one, but what if?

Also in the conclusions for those questions, do we usually say "not enough evidence for association" or "they are not independent", because I'm getting mixed messages from the mark schemes. Sometimes they seem to be sure like "Money has NO relation to happiness" and sometimes they're like "Oh, I guess we don't have enough evidence of a relationship between money and happiness". I'm like, bitch, MAKE! UP! YOUR! MIND!!


Combine them until the expected frequency is greater than 5, I've attached an example I did today

1463777170206.jpg

Usually I write there is not significant evidence to suggest that _____ is related to ______


Posted from TSR Mobile


lmao I did this today. A good check is that the sum of the d must be 0 and work from there :smile:

You write the coloumns and then rank them since they are not in numerical rank :smile:
Original post by L'Evil Wolf
was it about temperature and finishing position got caught out by it. I ranked one in the wrong order, taught me not too rush things, be wary.

Are solomon S3 easier than edexel papers? Difficulty about the same I think.


I did that question too, but I presume that the ranking order doesn't matter as it comes up as the same correlation in the negative sense resulting in exactly the same conclusion. I may ask an examiner who works at my school to confirm that though, just incase :smile:

Original post by L'Evil Wolf
lmao I did this today. A good check is that the sum of the d must be 0 and work from there :smile:

You write the coloumns and then rank them since they are not in numerical rank :smile:


Ah I hadn't realised about the sum of the differences! I'll use that in future to check

I didn't read the question and assumed I could go straight into it, lesson really to not skip words :tongue:
Original post by Euclidean
Combine them until the expected frequency is greater than 5, I've attached an example I did today

1463777170206.jpg

Usually I write there is not significant evidence to suggest that _____ is related to ______


Posted from TSR Mobile

Oh I'm actually talking about the one where we want to test associations of two things, not good fit tests. With good fit, you almost always get an expected frequency less than 5.
Original post by gagafacea1
Oh I'm actually talking about the one where we want to test associations of two things, not good fit tests. With good fit, you almost always get an expected frequency less than 5.


Oh sorry I totally misunderstood :redface:

I presume with contingency tables we can't merge it in a useful way so we have to just use it as is without merging?

I also suspect that they would deliberately avoid less than or equal to 5 just because of that reason
Original post by Euclidean
I did that question too, but I presume that the ranking order doesn't matter as it comes up as the same correlation in the negative sense resulting in exactly the same conclusion. I may ask an examiner who works at my school to confirm that though, just incase :smile:



Ah I hadn't realised about the sum of the differences! I'll use that in future to check

I didn't read the question and assumed I could go straight into it, lesson really to not skip words :tongue:


ah cool, I got negative, for that but I just marked it wrong haha. The H0 stuff was right though.

Yes that is key in S3, I did this french english qquestions on paper a or b and messed the 2 times longer part and put it on the other side of the inequality, I find it useful to compare my test statistic to what the actual data is to give me a true gauge of whether my answer is accurate or not.
Original post by L'Evil Wolf
ah cool, I got negative, for that but I just marked it wrong haha. The H0 stuff was right though.

Yes that is key in S3, I did this french english qquestions on paper a or b and messed the 2 times longer part and put it on the other side of the inequality, I find it useful to compare my test statistic to what the actual data is to give me a true gauge of whether my answer is accurate or not.


We have been doing the same questions today :lol:
You just doing solomon papers now?

I have only done A-C all day though, pretty abysmal
Original post by Euclidean
We have been doing the same questions today :lol:
You just doing solomon papers now?

I have only done A-C all day though, pretty abysmal


Kind of yes, although I have tackled some of the earlier papers.

I am saving papers D E F to be taken very soon Lol.


Oh that's okay, you just have to start with A, B, C ... in the top row and then have the corresponding ranks for the enthusiast and price in the two rows below. That's come up on edexcel past papers as well
Original post by L'Evil Wolf
Kind of yes, although I have tackled some of the earlier papers.

I am saving papers D E F to be taken very soon Lol.


Ah that's reassuring. I really wanna finish them all before wednesday but it's looking doubtful unless I get a move on :tongue:
Original post by Nikhilm
Oh that's okay, you just have to start with A, B, C ... in the top row and then have the corresponding ranks for the enthusiast and price in the two rows below. That's come up on edexcel past papers as well


I'll certainly be careful now

Completely skipped my mind when I did it
**** off Chapter 2. Go die in a well.


Posted from TSR Mobile
Original post by Euclidean
Ah that's reassuring. I really wanna finish them all before wednesday but it's looking doubtful unless I get a move on :tongue:


How did you get the confidence interval stuff, still having trouble.

Are solomon easier than edexcel do you think?
Original post by Euclidean
Oh sorry I totally misunderstood :redface:

I presume with contingency tables we can't merge it in a useful way so we have to just use it as is without merging?

I also suspect that they would deliberately avoid less than or equal to 5 just because of that reason


I guess you're right. Hoping to the gods they don't throw us a curve ball!
Original post by physicsmaths
**** off Chapter 2. Go die in a well.


Posted from TSR Mobile


There's no clear answer to half of them
Original post by physicsmaths
**** off Chapter 2. Go die in a well.


Posted from TSR Mobile


its not even that hard though and doesnt even cover much marks in exam.

But it belongs in Science and Statistics
Any really hard non-edexcel papers? (i.e. a specific solomon paper/AEI)
Original post by Euclidean
I'll certainly be careful now

Completely skipped my mind when I did it


Aha yeah like 5 marks down the drain

Quick Reply