The Student Room Group

A level maths hypothesis testing help understanding topic

Hi,

I've been studying hypothesis tesing in a level maths, but i'd just like some help wrapping my head around a paticular concept.

Lets say I think a coin is biased towards heads. So i flip the coin 10 times and i get 7 heads. Staight away i know that my null hypothesis is 0.5 and my alternative hypothesis is X>0.5 - but since i got 7 out of 10 heads, i used to think "well that's 7 out of 10, so the chance of me getting heads is 0.7, so reject null hypothesis" - however when i draw out the table below using binomial distribution, I can see that the chance of me getting 7 heads is 0.12 - not 0.7 like i thought.

Where's the error in my understanding?

Any help would be appreciated :smile:
(edited 11 months ago)
Reply 1
Original post by Muhammad012
Hi,

I've been studying hypothesis tesing in a level maths, but i'd just like some help wrapping my head around a paticular concept.

Lets say I think a coin is biased towards heads. So i flip the coin 10 times and i get 7 heads. Staight away i know that my null hypothesis is 0.5 and my alternative hypothesis is X>0.5 - but since i got 7 out of 10 heads, i used to think "well that's 7 out of 10, so the chance of me getting heads is 0.7, so reject null hypothesis" - however when i draw out the table below using binomial distribution, I can see that the chance of me getting 7 heads is 0.12 - not 0.7 like i thought.

Where's the error in my understanding?

Any help would be appreciated :smile:


Using your argument the probabillity of getting 10 heads out of 10 tosses is 1? Obviously its 0.5^10.
Reply 2
I'd also like to add that conceptionally, I know it's wrong. The probability would look like a bell curve - the chance of me getting 1 to 3 heads is really low, and then im very likely to get between 4 to 6 heads, and then im unlikely to get between 7 to 10 heads - but from gcse, I'm used to thinking that "since i got 7 heads, the probability is 0.7 (so quite high)" - like i said theres an error in my understanding here somewhere.
Reply 3
Original post by Muhammad012
I'd also like to add that conceptionally, I know it's wrong. The probability would look like a bell curve - the chance of me getting 1 to 3 heads is really low, and then im very likely to get between 4 to 6 heads, and then im unlikely to get between 7 to 10 heads - but from gcse, I'm used to thinking that "since i got 7 heads, the probability is 0.7 (so quite high)" - like i said theres an error in my understanding here somewhere.

The basic way to think of a binomial distribution is from a tree when the nCr counts the number of leaves and the p^n represents the probability of each branch. Your argument seems to be something like if I have 10 coins in a bag and 7 are heads, then the chance of picking one head in a single trial is 0.7. Binomails deal with multiple, independent trials, (tree branch) so you multiply the probabilities and count the number of leaves with the same outcome (nCr)

Quick Reply

Latest