Hi Sir, can you give a summary of how the continuity correction works please? I have watched the video about it but can't seem to wrap my head around it, thanks.
Loads of the mark schemes do this, but I don’t advise it as my preferred method - because calculators now can change the value of mu and sigma. They have the Z values as calculators previously didn’t do this, and you had to use the tables in the formula booklet.
Hi Sir, I have two questions if you don't mind answering. 1) would you have to use a contnuity correction every time we approximate the mean and standard deviation? 2) Could you clarify what you mean by part e at 8:40 since i am not too sure about it. Thanks for the great video!
You only need to use a continuity correction for things that start as binomial and end as normal - so not for finding sigma or mu! Part e feels a bit like a hypothesis test, right? The chance of what was observed occurring is incredibly low (lower than 5%), so this makes us think that the company's claim (which is like our H0) here, is not true - which is why I am 'rejecting' H0 here and saying their claim is not supported. I hope that helps!
For part e are we saying in other words that 132 seeds germinating should have happened according to the company’s claim , but 150 seeds germinated, the probability of 150 seeds germinating is very small, because its more than what the company claimed its different? would that make sense for an answer?
Yes that's right! I would probably say that we were expecting 132 seeds rather than it should have happened, as we know that there's a range of 'sensible' outcomes here.
In your previous video for approximating binomial p(x>9) was one of the discrete and to change it to contiinous you made is become p(x>9.5) so how comes @7:12 it isnt p(x>150.5) for this one?
So we're saying at least 150 - this means it could be 150 or higher. So to account that it could be 150, we go for 149.5, as that would still round to 150. 150.5 would round to 151, so would be for P(X>150), not P(X>=150), which is what we want!
Write out the info carefully, and decide what the distribution is. Once you’ve done that, it becomes easy! Take your time with the first stage, it’s the most important part.
Although it is good for the customer that more than 55% germinate, which is indicated by the low percentage of 1.16%, the question asks if the proportion of them germinating is different from 55%, which is true, as 1.16% is below the significance level. So it’s not about whether it is good or bad, but rather whether the claim is true or not. Hope that helps!
@@BicenMaths Right. I just wanted to be sure that I understood it right. That the company was under promising and overdelivering. Appreciate your response 😊
To calculate the variance for a normal approximation of a binomial, we use np(1-p), so 240x0.55x0.45 which is 59.4 for the variance. So standard deviation is the square root of this!
Hi Sir at 2:26 when you are doing the binomial distribution if you have a graphics calculator do we have do write out the process of 1- p(x=15) to get all the marks or can we skip it? Thanks
You can skip it - I definitely would! There was some guidance a while back about calculators where it basically said “if the students can do it on the calculator we should expect that some of them will do it on the calculator”. Good to hear you have the graphics, it takes a lot of work out of the binomial and normal distribution calculations!
Probably - as the normal distribution doesn't technically have an upper limit, whereas the binomial does. So if you're doing normal, and there's no upper limit, put in a big number! Sometimes it doesn't actually need to be that huge, as after 3 or 4 standard deviations away from the mean it really gets very very unlikely to occur. Hope that helps!
hi sir for 6:38 if we increase 150 by one, it becomes y is equal to or greater than 151. Then, if we scale it down by 0.5 it becomes y is equal to greter than 150.5. Why is this incorrect?
Hi Sir, can you give a summary of how the continuity correction works please? I have watched the video about it but can't seem to wrap my head around it, thanks.
1) change all questions from < or > to =
2) enlarge the range by 0.5 at each end (if 2 ends)
That’s it!
e.g P(34
Hi, for part d why did the mark scheme use z values to approximate a normal distribution?
Loads of the mark schemes do this, but I don’t advise it as my preferred method - because calculators now can change the value of mu and sigma. They have the Z values as calculators previously didn’t do this, and you had to use the tables in the formula booklet.
Hi Sir, I have two questions if you don't mind answering. 1) would you have to use a contnuity correction every time we approximate the mean and standard deviation? 2) Could you clarify what you mean by part e at 8:40 since i am not too sure about it.
Thanks for the great video!
You only need to use a continuity correction for things that start as binomial and end as normal - so not for finding sigma or mu!
Part e feels a bit like a hypothesis test, right? The chance of what was observed occurring is incredibly low (lower than 5%), so this makes us think that the company's claim (which is like our H0) here, is not true - which is why I am 'rejecting' H0 here and saying their claim is not supported. I hope that helps!
Do i always apply continuity correction when im switching a probability from binomial to normal?
Yes, always!
For part e are we saying in other words that 132 seeds germinating should have happened according to the company’s claim , but 150 seeds germinated, the probability of 150 seeds germinating is very small, because its more than what the company claimed its different? would that make sense for an answer?
Yes that's right! I would probably say that we were expecting 132 seeds rather than it should have happened, as we know that there's a range of 'sensible' outcomes here.
In your previous video for approximating binomial p(x>9) was one of the discrete and to change it to contiinous you made is become p(x>9.5) so how comes @7:12 it isnt p(x>150.5) for this one?
So we're saying at least 150 - this means it could be 150 or higher. So to account that it could be 150, we go for 149.5, as that would still round to 150. 150.5 would round to 151, so would be for P(X>150), not P(X>=150), which is what we want!
for part b) i get confused when there r 2 different things like the trays and seeds, what is the way to go about this?
Write out the info carefully, and decide what the distribution is. Once you’ve done that, it becomes easy! Take your time with the first stage, it’s the most important part.
Hi Sir, how do you know whether you're using binomial pd or binomial cd
PD is used for probabilities of X= and CD for probabilities of X
Hello Sir, On the hypothesis test that company claim in incorrect then is the percentage of germination > 55% ? (which therefore is a good thing ?)
Although it is good for the customer that more than 55% germinate, which is indicated by the low percentage of 1.16%, the question asks if the proportion of them germinating is different from 55%, which is true, as 1.16% is below the significance level. So it’s not about whether it is good or bad, but rather whether the claim is true or not. Hope that helps!
@@BicenMaths Right. I just wanted to be sure that I understood it right. That the company was under promising and overdelivering. Appreciate your response 😊
@@BicenMathsthey don’t give a significance level in the question, is it just assumed to be 5%?
@@rubiksworld2170 They should always give it - if they asked you to suggest a significance level (very unlikely) then 5% is typical
Sir at 6:26 why does y~N() have a sd of 59.4, i get why the mean is 132 as 0.55 is close to 0.5,but idk why the standard deviation is 59.4
To calculate the variance for a normal approximation of a binomial, we use np(1-p), so 240x0.55x0.45 which is 59.4 for the variance. So standard deviation is the square root of this!
Hi Sir at 2:26 when you are doing the binomial distribution if you have a graphics calculator do we have do write out the process of 1- p(x=15) to get all the marks or can we skip it?
Thanks
You can skip it - I definitely would! There was some guidance a while back about calculators where it basically said “if the students can do it on the calculator we should expect that some of them will do it on the calculator”. Good to hear you have the graphics, it takes a lot of work out of the binomial and normal distribution calculations!
@@BicenMaths okay thank you
Sir does it matter if for part d - if we said the upper is 240 (instead of a really big number)
Probably - as the normal distribution doesn't technically have an upper limit, whereas the binomial does. So if you're doing normal, and there's no upper limit, put in a big number! Sometimes it doesn't actually need to be that huge, as after 3 or 4 standard deviations away from the mean it really gets very very unlikely to occur. Hope that helps!
@@BicenMaths thanks a lot
hi sir for 6:38 if we increase 150 by one, it becomes y is equal to or greater than 151. Then, if we scale it down by 0.5 it becomes y is equal to greter than 150.5. Why is this incorrect?
At least 150 is Y>=150. You enlarge the range by 0.5 so it becomes 149.5
I’m bit confused - how come you’re using Y and not X?
im so confused, why is P(X>150) = P(Y>149.5)
Because of continuity corrections! P(X>=150) is discrete, P(Y>149.5) is continuous. There’s a video on this in the playlist!
I really don't understand part e even after watching this part 5 times
Is our answer to d suggesting that the company's actual percentage of seeds that germinate should be above 55%?
Yes, that’s absolutely right - 150/240 seeds is a higher proportion than 55%, and the low percentage calculated supports this too.
@@BicenMaths oh OK so we are saying because the percentage is so low , it couldn't have happened by chance